ISSN: 2089-208X
Hee
ANNUAL INTERNATIONAL CONFERENCE
SYIAHKUALAUNIVERSITY (AIC UNSYIAH)
CHAPTER
‘AAC Dayan Dawood, Darussalam - Banda Aceh, Indonesia, November 29-30, 2011
PROCEEDINGS
@()@o
Editors:
Hesti Meilina, indonesia
Hizir, Indonesia
Abdul Razak Salleh, Malaysia
Amir T.P. Najafabadi, Iran
Mohammad Jawaid, India
Sun NING ea eee
mandirEDITORIAL BOARD
Editor-In-Chief
Dr. Hesti Meilina (Environmental Engineering, Indonesia)
Editorial Members:
Dr. Hizir Sofyan (Mathematics, Indonesia)
Dr, Abdul Razak Salleh (Mathematic, Malaysia)
Dr. Amir T-P. Najafabadi (Mathematic, Iran)
Dr, Mohammad Jawaid (Engineering, India)
Organizer:
Syiah Kuala University Research Institute
Publisher:
Syiah Kuala University Press
Kopelma Darusalam - Banda Aceh 2311, Indonesia
Cover Design by. Syobobudin STNOTICE OF DISCLAIMER
The authors and publisher of these proceedings have used their best efforts in
preparing and collecting the materials. These efforts include the development, research and
testing of these theories to determine their effectiveness. However, the publisher bears no
responsibility for any damage resulting from any inadvertent omission or inaccuracy in the
proceedings. Further, the authors and publisher make no warranty of any kind, expressed
or Implied, with regard to the documentation contained in these proceedings. The authors
of papers published in these proceedings have transferred their copyright to Syiah Kuala
University. The opinions expressed by the authors do not necessarily reflect those of Syiah
Kuala University and are solely the responsibility of the authors.
‘All manuscripts printed in these proceedings were reproduced directly from the
softcopies submitted by the authors (unedited), with only minor editorial amendments.
Although all efforts were made to ensure the accuracy and correctness of the format, the
Committee of Annual International Conference Syiah Kuala University will not be
responsible or liable for any unintentional errors made during the publication or any conflict
of interest of theis proceedings.
Individual readers of these proceedings and nonprofit libraries acting for them are
freely permitted to make fair use of the material, such as to copy a paper for use in
teaching or research. Permission is granted by Syiah Kuala University Press to quote from
these proceedings with the customary acknowledgment of the source. Republication or
systematic or multiple reproduction of any material in these proceedings Is permitted only
with written permission from Syiah Kuala University Press,
Request to republish or reproduce papers from these proceedings should be
addressed to Syiah Kuala University Press, Kopelma Darussalam, Banda Aceh 23111,
IndonesiaMESSAGE FROM THE RECTOR
On behalf of Syiah Kuala University (Unsyiah), I would like to
extend my warmest welcome to all participants to the Annual
International Conference of Syiah Kuala University 2011 (AIC-
UNSYIAH 2011).
Syiah Kuala University is the largest and the oldest national
university in Aceh Province, Sumatera Island, Indonesia. The
university was established on June 21st, 1961, The establishment
of UNSYIAH was driven by a Spirit to farm an institution to bring
the Acehnese to become educated, knowledgeable and pious to
God Almighty.
UNSYIAH has a vision to establish itself as an innovative,
independent, and outstanding university in terms of the
development of science, technology, humanities, sport and arts, in order to produce
qualified graduates who highly honor moral and ethical values.
This year, UNSYIAH has hosted a series of events commemorating its 50th anniversary. As
part of the celebration, the university held an interdisciplinary academic conference
between of November 29-30, 2011. The conference included plenary addresses, oral &
poster parallel sessions and a tsunami/historical sites tour.
T would like to thank our keynote and invited speakers for their time and support for this
conference. I am excited to witness such development in the research scene,
Lastly, I would like to thank the Organizing Committee for their hard work in making this
event successful. To all participants, I wish you fruitful interactions with your peers and our
foreign friends as well as, an enjoyable stay in Banda Aceh.
Thank you:
Sincerely
Professor Dr. Darni Daud, M.Ed
Rector of Syiah Kuala UniversityMESSAGE FROM THE CHAIRMAN
1 would like to take an opportunity to warmly welcome all the
participants of the Annual International Conference of Syiah
Kuala University 2011 (AIC- UNSYIAH 2011) to Banda Aceh,
Indonesia.
This conference was organized by The Syiah Kuala University
Research Institute, The institute is honored to host this
distinguished and meaningful event. This conference signifies
yet another achievement by the Research Centre in bringing
together scholars from all over the world to promote scholastic
work. It is also in line with the University’s commitment to the
advancement of knowledge through pursuing scientific
discoveries and technological innovation.
The response for the conference has been overwhelming. We are glad to have received
abstract submissions and approximately 180 papers from 9 countries, however due to
limitation of time slots only about 107 papers will be able to be presented. Among the
participants, there are not only experienced experts and researchers but also postgraduate
students from emerging areas such as agriculture, marine and fisheries, biology and
chemistry in general, health sciences, law and politic, education, history and economics.
The presented papers have been published in three different ‘chapters; Life Sciences,
Sciences and Engineering, and Education and Social Sciences.
We hope that participants will exploit the great opportunity provided by this conference to
link with some of the world’s best minds and establish collaborative ties. We also hope that
it will spur inspirations for more innovative ideas. We look forward to welcoming you in the
Annual International Conference of Syiah Kuala University 2011 and wish you a pleasant
and enjoyable stay in Banda Aceh.
Thank you
Sincerely
Dr, Musri Musman, M.Sc
Chairman/Director of Syiah Kuala University Research InstituteCONTENTS
EDITORIAL BOARD
NOTICE OF DISCLAIMER
MESSAGE FROM THE RECTOR
MESSAGE FROM THE CHAIRMAN
CONTENTS
SCIENCES AND ENGINEERING SESSION
1
10
1
12
13
14
15
16
Preparations of Chemical Sensors for Simple Formalin Detection
in Contaminated Food
‘Adlirn,Hasan,zartaida Fitri, Yulida Amri,Martina Sari,and Saiful Mahya
Microbial Community Analysis during Start-up of Anaerobic Co-digestion
based on Quinone Profiles using Supercritical Fluid Extraction
sri Gani,Ahmed Fall, and Hiroyuki Daimon
Screening of Shallow Groundwater in Aceh Besar and
Banda Aceh Districts for Contamination with Heavy Metals
. YulvizaR,
Comparative Performance Study of Two Simple Soot Models for the
Prediction of Soot Level in Atmospheric Turbulent Non-Premixed Flames
Elwina, Yunardi, Novi Sylvia, Wusnah, and Yazid Bindar
A Study of Hydrogen Storage in Zirconium-containing NaY5.7 Zeolite
Maulana
Photocatalytic of Hydrogen Production from Water over Ni/NIO core/shell
Nanoparticle-loaded La0.02Na0.98Ta03
‘Husni Husin,Bing-Joe Hwang, and Wei-Nien Su
Preliminary Study on the Adsorption of Lead (II) Ions from
Aqueous Solution with Breadfruid’s Bark (Artocarpus altilis) by
Un-modified and Modified with Citric Acid
Lia Mairiza, Muhammad Zaki, Nurhayati, and Evi Jullyanti
Analysis of Turbulence Models Performance for the Predictions of Flow Yield,
Efficiency, and Pressure Drop of a Gas-solid Cyclone Separator
‘Novi Sylvia, Yunardi, tiham Maulana, Elwina, Wusnah and Yazid Bindar
Preparation of Asymetric Cellulose Acetate Membrane by Phase Inversion
Method of Cellulose Pulp from Wood of Sengon (Paraserianthes falcataria)
‘cm. Rosnelly
Phenol Degradation on Heterogeneous Catalytic Oxidation by using
Cobalt-Natural Zeolite Catalyst
Syaifullah Muhammad, Edy Saputra, Shaobin Wang, and Moses 0. Tadé
Computational Fluid Dynamics of Crosswind Effect on a Flare Flame
Wusnah, Yunardi, Ilham Maulana, Elwina, Novi Sylvia and Yazid Bindar
Optimization of Cyclone Geometry for Maximum Collection Efficiency
Yunardi, Ilham Maulana, Elwina, Wusnah, Novi Sylvia and Yazid Bindar
Oil Yield of Green Microalgae Isolated from Ponds around Banda Aceh City
‘Marwan, Suhendrayatna, and Bahagia
Relationship between Personality and Coping with Stress among
Undergraduates of Universiti Kebangsaan Malaysia
Choong-Yeun Liong and Hui-Ping Yeoh
Malay Mathematics based on Terengganu Inscription
Abdul Razak Salleh
Exact Solutions for A Class of Matrix Riemann-Hilbert Problems
‘Amir T. Payandeh Najafabadi and Kucerovsky
Page
16
25
32
38
46
53
et
66
75
80
86
o1
99
1127
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Applying Genetic Algorithm on Power System Stabilizer for Stabilization of
Power System
‘Arnawan Hasibuan and Syadrudin
Effect of Cavitation in Cylindrical and Two-Dimensional Nozzles on Liquid Jet
Formation
‘Muhammad Iiham Maulana and Jalaluddin
Geoelectrical Characterization for Liquefaction at Coastal Zone in South Aceh
‘Muhammad Syukri
‘Analysis Quality Control (QC) on CT Scan in RSUZA Banda Aceh an Effort
to Get the Best Quality in Image
SaftriR. Evi ¥, Heru P, Bustami
Planning and Analyzing DVB-T Technology in Pulo Aceh using High Altitude
Platform Station
Rizal Munadi, Anwar Sidi, Roslidar
Simulation of Sea Surface Temperature (SST) and Sea Surface Salinity (SSS)
in the Bay of Bengal
‘Syamsul Rizal, Muhammad, Taufiq Iskandar, Ichsan Setiawan,
‘Agus Satriod, and Radinal!
Comparison Study Towards an Old Acehnese Fara‘idh Table
Solmawaty ari
Early Detection of Rotor-Bar Faults of Three-phase Induction Motor Using Motor
Current Signature Analysis Method
Syatrudin and Arnawan Hasibuan
The Performance of Asphalt-Concrete Wearing-Course (AC-WC) Mixture by
Using Rice Husk Ash as Filler with the Addition of Asbuton In
Asphalt Pen 60/70 as Binder
‘Maizal Kamil, Reni Anggraini,Fitrika Mita Suryani
Aphorpines and Benzylisoquinoline Alkaloids from Bark of
Cryptocarya crassinervia
‘Nurdin Saidi
‘A Need for Affordable Housing in Banda Aceh, Indonesia
‘Nova Iriansyah
Isolation and Crystallization of Patchouly Alcohol. from Patchouly il
Yullant alsyah
Proposed long period transition map for new Indonesia earthquake
resistant building code based on Indonesia seismic hazard map 2010
Usama Juniansyah Fauzi, Achmad Fauzi, Masyhur Irsyam, FX. Toha
and’ Hendriyawan
Frequency of Use Minute concentrations of Cadmium in Aqueous Solution
by Near Infrared Spectroscopy and Aquaphotomics
Mollina H, Putra A. and Taenkova R.
Acetosolv Pulping Modeling of Oil Palm Frond Fibers
‘Nasrullah RCL and I. Mazian
122
127
134
140
145
153
160
165
174
184
190
195
202
210
218Proceeding of Annual international Conference Syiah Kuala University 2011
‘Banda Aceh, Indonesia. November 29-30, 2011
Analysis of turbulence models performance for the
predictions of flow yield, efficiency, and pressure
drop of a gas-solid cyclone separator
‘Novi Sylvia, "Yunardi, “Ilham Maulana, “Elwina, 'Wusnah and °Yazid Bindar
2 Chemical Engineering Dept., University of Malikussaleh, Lhokseumawe, Indonesia;
2chemical Engineering Dept., Syiah Kuala University, Banda Aceh, Indonesia; * Mechanical
Engineering Dept., Syiah Kuala University, Banda Aceh, Indonesia; * Chemical Engineering
Dept., State Polytechnics of Lhokseumawe, Lhokseumawe, Indonesia; ° Energy and
Processing System of Chemical Engineering Dept., Faculty of Industrial Technology,
Bandung Institute of Technology, Bandung, Indonesia. Corresponding Author:
yunardi@unsyiah.ac.id
‘Abstract. This paper presents the results obtained from the application of computational flud dynamics (CFD) to
‘modelling the flow felé of a Lapole cyclone anc to optimizing the cyclone based upon its geometrical parameters.
[A pre-processor software GAMBIT was employed to set up the configuration, discretisation, and boundary
contltions of the cyclone. ‘The characterises ofthe cyclone being studiad was 0.2 m in diameter, receiving a gas
flow rate of 0.1 m3/s with a particle mass loacing of 0.01 kg/m. A commercial CFD code FLUENT 6.2.16 was
lemployed to simulate the flow field and particle dynamics in the cyclone. The objectve of this research was to
Investigate the performance of a number of turbulence models on the prediction of the flow field, collection
eificlency ane pressure drop In the Lapple cyclone, & number of five turbulence models under Reynolds Averaged
Navier Stoxes (RANS) category, inclusing Spallar-Allmaras, standard k-e model, RNG ke model, standard k-w
‘model, ané Reynolds Svess Model (RSM) were examined In the simulation of the flow feld and particle cynamics
Inside’the cyclone. A validation of all calculation was performed by comparing the predicted results in terms of
axial and tangential velocites, efficiency and pressure drop against experimental data of a Lapple cyclone taken
from literature, The results of the investigation show that aut of five turbulence models being tested, the RSM
presented the best precited results. The pracictions of axial and tangential velocities as well as cyclone efficiency
by this model are in excelent agreement with the experimental data. Although the pressure drop in the cyclone Is
Lunder-predictad, the RSM predictions are far better than those of other model. Other turbulence models are over-
Predicted and under-predicted the axial and tangential velocty, respectvely. With respect to efficiency and
Pressure drop of the cyclone, other models are capable of fellowing the trend of the experimental data but they
fallec to agree with the experimental values. ‘These results suggest that the RSM Is the most sultable turbulence
‘model to represent the flow field and particle dynamics side a cyclone gas-solid separator,
Key words: cyclone, computational fluid éynamics, turbulence model, efflency, pressure érop
Introduction
Cyclones are one of the most common equipment used for controlling dust emissions of
gaseous flow in industrial processes. Although current engineering developments have
enabled to employ cyclone for example as dryers and reactors, their main application
remains in the area of air pollution control where high efficiencies are required to meet the
stringent regulations. In comparison with other equipment used for air pollution control,
cyclones are more preferable due to their simplicity in the design, inexpensiveness to
manufacture, low maintenance costs, and adaptability to a wide range of operating
conditions such as high temperature and pressure. Despite they are frequently used as
final collectors where large particles to be removed, it has been also a common practice to
employ cyclones as pre-cleaners for a more efficient collector such as an electrostatic
precipitator, scrubber or fabric filter (Swamee, 2009)
There are four major parts to a cyclone, the inlet, the cyclone body, the dust
discharge system, and the outlet al.| affect the overall efficiency of the cyclone. The
principle of cyclone separation Is simple: the flow of gas-solid mixture Is directed into the
cyclone through the inlet on the top section. Then, the cylindrical body induces a spinning,
53Proceeding of Annual international Conference Syiah Kuala University 2011
‘Banda Aceh, Indonesia. November 29-30, 2011
forcing particulate matter to the wall of cylinder. The gas continues down the cyclone body
to the cone, which gives the gas enough rotational velocity to keep the particulates against
the wall. At the bottom of the cone, the gas changes direction from downward to upward
‘The ascending vortex enters a tube extension that is sometimes called a vortex finder and
exits the cyclone. Meanwhile, the collected particulate matter drops into a hopper, where it
is periodically or continuously removed.
The main performance of a cyclone is primarily judged from its collection efficiency
and pressure drop. In spite the fact that its design and operation are simple, the flow
behavior and particle dynamics inside the cyclone are complicated, requiring efficient
mathematical models to provide accurate predictions of the efficiency and pressure drop for
the purpose of design and operation of a cyclone. Numerous semi-empirical models leading
to the predictions of collection efficiency and pressure drop have been developed by many
investigators. Leith (1990) summarized a number of these models, including those
developed by Stairmand (1951), Barth (1956), Shepherd and Lapple (1939), Lapple
(1951), Leith and Licht (1972). Although the majority of the empirical models were
developed based on the experimental data of particular cyclone geometry, a substantial
error between the prediction and measured data in the cyclone efficiency is evidence due
the use of different assumptions and geometry conditions. As a consequence, the use of
semi-empirical models has limitation in the prediction of cyclone performance. Numerical
methods are therefore proposed to model the flow field and particle dynamics of these
devices for the purpose of predicting the collection efficiency and pressure drop.
‘After the first numerical simulation of cyclone using computational fluid dynamics
(CFD) technique (Boysan et al., 1982), a number of CFD modeling works on the cyclone
performances have been performed. Elsayed and Lacor (2010) optimized the cyclone
geometry aiming at obtaining minimum pressure drop with the aid of response surface
methodology. They used Reynolds Stress Model to represent the flow field inside the
cyclone and the results showed that the most significant geometrical parameters are the
vortex finder diameter, the inlet section width, the inlet section height and the cyclone total
height. Zhou and Soo (1990) applied two-equation turbulence model, k-e standard for the
modeling of gas solid flow and collection of solid in a cyclone separator. They found out
that the k-e standard Is capable of providing good predictions with respect to axial velocity;
however, it fails to simulate strongly swirling flow near the axis. Hoekstra et al. (1990)
evaluated the performance of the k-e model, the RNG-k-e model, and the Reynolds stress
transport model (RSTM) in predicting the gas flow field in a cyclone separator. They found
out that the RSTM demonstrated reasonable predictions in terms both axial and tangential
velocities inside the cyclone. On the contrary, the other two models were considered not
sultable for the predictions of cyclonic flow. However, it is important to note here that the
RNG-k-e model could predict well the pressure drop in three type of cyclone sampler
(Griffiths and Boysan, 1996). This finding suggests that any simpler turbulence model does
not always results in poor prediction, as also demonstrated by Suyitno (2005) when one-
equation turbulence model, Spallart-Allmaras, was found to give better predictions than
that of the RNG-k-e model in a study of cyclone performance. On this basis, it is necessary
to investigate the current commonly used turbulence models in an attempt to have a better
understanding on the performance of each model in the simulation of flow and pressure
fields as well as efficiency of a cyclone.
‘The objective of this paper is to present the results of evaluation of the performance
of five turbulence models under Reynolds Averaged Navier Stokes (RANS) category, namely
Spallart-Allmaras, standard k-e model, RNG k-e model, standard k-w model, and Reynolds
Stress Model (RSM) in predicting flow and pressure fields as well as efficiency in a Lapple
cyclone separator. Validations of predictions are made by comparing the predictions
resulted by each model with experimental data taken from Wang et al. (2005).
54Proceeding of Annual international Conference Syiah Kuala University 2011
‘Banda Aceh, Indonesia. November 29-30, 2011
Methodology
Turbulence Models Description
Turbulence theories, simulation and modelling have always been important subjects in fluid
dynamics and engineering, descriptions of different turbulence approaches can be found in
various computational fluid dynamics textbooks, Any modelling technique involves a
number of descriptive equations whose solution needs to be obtained numerically. In
general, with regards to turbulence prediction alone, three main classes of numerical
simulations are currently being developed: (i) direct numerical simulation (DNS); (ii) large
eddy simulation (LES); and (ill) Reynolds averaged Navier-Stokes (RANS) approaches.
‘The DNS of turbulent flows essentially involves a full numerical solution of the time
dependent Navier-Stokes equations and accommodates all time and length scales of
turbulence. From the conceptual point of view, it is fundamentally the simplest method to
implement, since no turbulence modelling is needed. In DNS, all of the turbulent motions
are resolved in the computational model from the largest scale to the smallest scale of
turbulent eddy. As a consequence, the computational domain should be large enough to
contain the largest eddies, and the grid spacing should be fine enough to resolve the
smallest eddies. Therefore, it is extremely expensive to simulate even the simplest types of
flow (e.g. homogeneous turbulence), primarily due to the refined grid required to resolve
the small-scale turbulence structures, as well as the small time-steps required for the time-
scales of the smallest eddies. In the Reynolds-averaged Navier- Stokes (RANS) approach,
instead of directly solving for the turbulence field, solutions are obtained by solving time-
averaged transport equations.
‘The approach models all scales and solves the governing time-averaged equations
which introduce unknown apparent stresses known as the Reynolds stresses. This adds
a second-order tensor of unknowns for which various models can provide different levels
of closure. Basically, two distinct types of RANS model have been developed: first-
moment closure models and second- moment closure models. In the former, the
unknown Reynolds stresses are reduced by correlation with the first-moment. The second
moment closure models approximate the higher-order moments (ie. the triple
fluctuating velocity correlations) by second- moment terms, and solve transport
equations for the Reynolds stresses directly. As a consequence of modeling the unknown
terms, RANS turbulence models like Spalart-Allmaras, standard k-e model, RNG k-e model,
standard k-w model, and Reynolds Stress Model (RSM) are capable of producing much
faster computation in comparison to those of LES and DNS. This is the reason to use such
models in the present investigation.
Numerical Computation
All transport equations presented in the turbulence model description above are numerically
solved using @ commercial CFD code, Fluent 6.2.16 (Fluent Inc., 2005) . Control volume
approach was used to discretize the transport equations. The SIMPLE algorithm was used
to solve pressure-velocity coupling and first-order and second-order interpolation schemes
for turbulent kinetic energy and momentum equations, respectively. Flows inside the
cyclone was assumed to be in steady state. The numerical computation was carried out
with an accuracy of 10-3 for the entire flow field parameters.
Computational Domain and Boundary Conditions
Figure 1 on the left illustrates a Lapple cyclone, having a diameter of 0.2 m employed in
this study with the ratio of geometric parameters is shown in Table 1. On the right of
Figure 1, the cyclone geometry drawn using GAMBIT code was set up with boundary
conditions. Initially, the geometry of the cyclones has @ number of 57,000 cells. However,
during computation in FLUENT, a mesh adaptation was performed allowing the increase of
the number of cells of around 30% of the initial state, Information on material data for the
cyclone computation is presented in Table 2. It should be noted here that these data are
similar to the experimental data presented by Wang et al. (2005).
55Proceeding of Annual international Conference Syiah Kuala University 2011
‘Banda Aceh, Indonesia. November 29-30, 2011
Table 1. Cyclone geometry used in this study (D=0.2 m)
ao _b/D De/D__S/D. nD HD B/D
02505 oS 0625 2.0 a0 0.25
Table 2. Material data used as input of the cyclone calculation
Temperature of air flow 25°C
Min. diameter of particle Sum
Max. diameter of particle 200 um
Mean diameter of 29.90 um
particle
Spread parameter 0.806
Ash density 3320 kg/m?
Air densi 1.225 kg/m?
a wut Ott
Figure 1. Cyclone geometry used in this study
Flow Field Calculation and Validation
Flow field computation was carried out using FLUENT 6.3 with varying turbulence model as
listed in Table 3. All constant used in each model were those of default without any
adjustment. Computation were performed using a laptop of 2.00 GB RAM having a 2.13
GHz of speed, The computation time for each run highly depended on the turbulence model
used. However, in general each run required a computation time from 4 to 24 hours.
Table 3. Turbulence mode! were tested in this study
Classification of model Derivative Model Tested model
‘One equation model__Spalart Allmaras
1. ke stander
RANS Two equation model 2- RNG ke
3. komega
Reynold Stress Model RsW@_——
56Proceeding of Annual international Conference Syiah Kuala University 2011
‘Banda Aceh, Indonesia. November 29-30, 2011
‘The validation of the predictions was performed by comparing with the experimental
data reported by Wang et al. (2005). In this experiment, Wang et al. (2005) used a Lapple
cyclone model, Experimental measurements were presented in the forms of flow
tangential and axial velocities, pressure drop and efficiency.
Results and Discussion
Prediction and Validation of Flow Field in Cyclones
Figure 2 shows comparison of predicted axial velocity by various turbulence models and
experimental data, Both experimental data and predictions show that the axial velocity
reached the peak in the center of the cyclone having the lowest values at the positions near
the wall of the cyclone. All turbulence models qualitatively produce the similar trend,
following the evolution of experimental data, However, among five turbulence models
tested in this study, Reynolds Stress turbulence model (RSTM) produced more satisfactory
prediction. The RSTM Predicted axial velocity both in the area near the wall and in the core
of the cyclone was in close agreement with experimental data. Both qualitatively and
quantitatively, the performance shown by the RSTM in the prediction of axial velocity in
cyclone is more superior tan those of other turbulence models tested in this study. Although
the predictions of other models follow the trend of experimental data, turbulence models
such as the k-e standard, RNG k-e, k-w and Spalart Allmaras produced a substantial
Geviation in peak area. The main reason for the accuracy of the RSTM predictions is most
likely due to its performance to predict a complex flow involving swirling and vortex as in
the case of cyclone.
axial velocity, mis
radial position, m
Figure 2. Comparison between axial velocity predictions with experimental data (symbols =
experimental data; Line = Prediction results; ; — RSM; — — k-e; + k-e RNG; —
= k-w standar; ~~ - Spalart Allmaras)
Figure 3 shows a comparison of predicted tangential velocity by various turbulence
models and experimental data. Tangential velocity profile along the radial position inside
the cyclone is different from the axial velocity profile. The axial velocity forms a single peak
at the core of the cyclone, while the tangential velocity profile forms two peaks in the left
and right of the centerline, with the minimum peak occurs in the core of the cyclone. With
regard to the prediction of tangential velocity by various turbulence models, all turbulence
models quantitatively were unable to capture the evolution of the experimental data
Although qualitatively, predictions by RSTM is superior than those of other turbulence
models, it has not been able to capture the tendency of the tangential velocity data as its
ability to capture the trend of the axial velocity data. It should be noted here that all
models were run on the basis of the default of Fluent code, without adjusting any constant
57Proceeding of Annual international Conference Syiah Kuala University 2011
‘Banda Aceh, Indonesia. November 29-30, 2011
in the model. Therefore, the improvement of the prediction, particularly by the RSTM could
be obtained by adjusting the empirical constants associated with spreading rate.
Tangential velocity, m/s
“010” 005000005 Oto
radial position, m
Figure 3. Comparison between the predictions of tangential velocity with experimental data
(symbols = experimental data; Line = Prediction results; ~ RSM} ~~ k-e; + ke
ERNG; + — k-w standar; ~ “'~ Spalart Allmaras)
Figure 4 illustrated axial velocity and tangential velocity contours in the left and
right, respectively in the cyclone. Inspection the left figure, it is clear that low axial velocity
dominated in the region closer to the wall. As the axial velocities have minus value,
meaning the direction of the flow is downward. This means that the majority of the
Particles are moving downward, although small percentage of particles are moving upward
as indicated by positive values of axial velocity in the core of the cyclone. Turning to the
right figure, the tangential velocities of between 13 and 17 m/s dominated the flow field in
the cyclone. It is this velocity that forced particles towards the wall went spinning
downward. In contrast to the left figure, the tangential velocity at the center-line has very
low values, indicating that the tangential velocity dominates the flow field at outside of the
core regions of the cyclone, while the axial velocity dominates the flow field in the core of
the cyclone.
Figure 4. Axial velocity (left) and tangential velacity contours (right) in the cyclone
58Proceeding of Annual international Conference Syiah Kuala University 2011
‘Banda Aceh, Indonesia. November 29-30, 2011
Prediction and Validation of the Efficiency and Pressure Drop
Figure 5 shows a comparison of efficiency prediction by various turbulence models against
experimental data on the basis of gas entrance velocity into the cyclone. Predictions
generated by RSTM is much better than those of other predictions, as they are closer with
the experimental data. Efficiency predictions by other turbulence models are mostly below
the experimental data, with the predictions generated by standar k-w represents the most
unsatisfactory. This is understandable as k-« model is not prepared to predict swirling flow
as occurs in the cyclone.
Eticiency
Velocity, m/s
Figure 5. Comparison between efficiency predictions and experimental data (symbols =
experimental data; Line = Results predictions— RSM; ~~ k-e; + k-e RNG; —
— k-a standar; ~ «+ ~ Spalart Allmaras)
a3
Pressure Drop (Pa)
eseRR 88
823298
he he
Velocity, mis
lure 6. Comparison between pressure drop predictions and experimental data (symbols
= experimental data; Line = Results predictions; — RSM; - ~ k-e; -- k-e RNG;
= += kew standar; ~ ++ — Spalart Allmaras)
Figure 6 presents a comparison of pressure drop predictions by various models of
turbulence with experimental data. The results shown that there was no significant
differences among predictions of one turbulence model to others, both qualitatively and
quantitatively, with the exception to the predictions generated by RSTM. All models
qualitatively produced a similar trend to the experimental data. However, quantitatively, all
59Proceeding of Annual international Conference Syiah Kuala University 2011
‘Banda Aceh, Indonesia. November 29-30, 2011
models could not provide a satisfactory performance when compared to the experimental
data. From all turbulence models tested, only the RSTM that produced better predictions
than those of others, although they under-predicted experimental data
Conclusions
From the results and discussions, the following conclusions are drawn:
1. Predictions of flow fields are significantly influenced by the performance of a
turbulence model to represent the vortex and swirling flows that occur in the
cyclone. Among five turbulence models being tested in this study, only the RSM
turbulence model is capable of producing reliable predictions of the flow field since
the model was prepared to predict a complex flow,
2. Results of the simulation shows that the tangential velocity governs the flow field
outside of the center line of the cyclone, forcing the particle towards to the wall
and causing the particles to fall towards to the cones region for collection, On the
other hand, the axial velocity dominates the flow field in the center line of cyclones,
especially closer to the outlet pipe, even though the values are smaller than those of
tangential velocity.
3. Better flow field predictions by RSM model provides positive impact on the
predictions of collection efficiency and pressure drop that occurs in the cyclone.
Calculated efficiency by the standard k-e, RNG ke, standard k-w and Spalart
Allmaras models were under-predicted, but the predictions generated by the RSM
model are in closer agreement to the experimental data.
References
Barth, W. 1956, Design and layout of the cyclone separator on the basis of new investigations,
Brenn.Warme Kraft, 8 :1-9.
Boysan, F., Ayers, W.H., Swithenbank, J.A, 1982, Fundamental mathematical modeling
‘approach to cyclone design. Trans IChemE, 60: 222-230.
Elsayed, K., Lacor, C, 2010. Optimization of the cyclone separator geometry for minimum
pressure drop using mathematical models and CFD simulations. Chem Eng Sci, 65: 6048-
6058,
Fluent 6.2.16. 2005. User's Guide. FLUENT Incoparated.
Griffiths, W.D, Boysan F, 1996, Computational fluid dynamics (CFD) and empirical modelling of
the Performance of a number of cyclone samplers. J. Aerosol Sci, 27: 281 -304
Hoekstra, A. J., Derksen, J. J., Van den Akker, H. E. A. 1999. An experimental and numerical
study of turbulent swirling fow in gas cyclones.Chem Eng Sci, 54: 2055-2065,
Lapple, C.E. 1951. Process uses many collector types. Chem.Eng, 58:144-151
Leith, D., Licht, W. 1972. The collection efficiency of cyclone type particle collectors. AIChE
Symp, 68: 196-206.
Leith, D. 1990. Cyclone performance and optimization. Technical Report, Department of
Environmental Sciences and Engineering, University of North Carolina.Chapel Hill, USA.
Stalrmand, C. J. 1951. The design and performance of cyclone separators. Trans IChemE, 29:
356-383,
Shepherd, C.B., Lapple, C.E. 1939. Flow pattern and pressure drop in cyclone dust collectors.
Industrial and Engineering Chemistry. 31(8): 972-984
‘Swamee, P.K., Aggarwal, N., Bhobhiya, K. 2009. Optimum design of cyclone separator. AICHE,
55: 2279-2283.
Suyitno. 2005. CFD of analysis Performance cyclone with Spalart-Allmaras model and RNG k- e.
Media Mesin, 6:47-54,
Wang, B., Xu, D.L., Xiao, G. X., dan Yu, A.B. 2005. Numerical study of gas-solid flow in a
Siklon. Appl Math Model, 30:1326-1342,
Zhou, L.X., S00, S.L. 1990.’ Gas-solids flow and collection of solids in a cyclone separator.
Powder Technol, 63: 45-53,
60Proceeding of Annual international Conference Syiah Kuala University 2011
‘Banda Aceh, Indonesia. November 29-30, 2011
Preparation of asymetric cellulose acetate
membrane by phase inversion method of cellulose
pulp from wood of sengon (Paraserianthes
falcataria)
C. M. Rosnelly
Department of Chemical Engineering, Syiah Kuala University, Banda Aceh 23111,
Indonesia. Corresponding author: cutnelly@ymail.com
Abstract: Cellulose acetate ultrafiltration membrane is one of kinds process of membrane which has been utlized
in separation and purification processes. Cellulose acetate ean be abtained from acetylation of cellulose sengon
wood (Paraserlanthes falcatara). Wood Sengon was used because It has potential as @ raw material, There are
three steps of cellulose clacetate manufacturing process: (1) activation of cellulose using acetic ace, (2)
acetylation using acetic anhydride as reactant and sulfuric acid as catalyst, (3) hydrolysis. At each operating
Process was observed at 50°C. Preparation of cellulose diacetate membrane using phase Inversion methode by
edition of cellulose ciacetate as polymer; N,N-Dimethy! formamide (DMF) as solvent, water as non-solvent, and
Polyethylene alycal (PEG). Membrane pore size was determined by measuring Molecular Weight Cut Of (MWCO)
Using Dextran and Bovin Serum Alburine (BSA) as standard solution. Membrane morphology was observed by
Scanning Electrone Microscope (SEM). Cellulose acetate of acetyl content of 39.66% and number average
‘molecular weight 130,221 Da was obtained 30 minutes of activation time; one hour acetylation, and 15 hours
hydrolysis. The addition of PEG procuced a thickner layer and suppress the formation of macrovold. SEM analysis
shows a denser stucture membrane morphology with better regularity of pore shape. The resulting membranes
twin were coagulated at lower temperature snaws dense structure. High fk with low rejection obtained from the
membrane wit greater porosity and pore distribution. MWCO determination was based on the value of 80%
rejection of dextran and BSA standard Solution and the obtained pore size ranges obtained 67 kDa, and is stl
categorized as ultrafiltration membrane,
Key words: Paraserianthes falcataia, cellulose diacetate inversion, cellulse diacetate membrane, utrafitration,
Introduction
Membrane is semipermeable barriers that separates two phases and restrict the transport
of various substances in a specific way (Strathmann, 1990). In recent years, membranes
and technology of membranes have grown from a labory scale. To day, technology of
membranes are used on a large scale such as medical care/ medical industry (hemodialysis,
purification of enzymes, antibiotics, etc), food industry (concentration of fruit juice, sugar
purification, waste water treatment, etc), other industries (desalination and salt production,
waste water treatment dan recovery of valuable, etc.) (Shibata, 2004)
Technology of membrane more efficient and economical because the processes are
faster, low energy consumption, operation at mild temperatures, the non-addition of
chemical products (Mulder, 1996; Countinho, 2009). In Indonesia, utilization technology of
membrane is not to develop in good because the materials of membrane is very difficult to
find, Cellulose acetate suitable as membrane materials because is an environmental friendly
substance, Cellulose acetate can be found from acetylation process of cellulose, In
Indonesia, Sengon plant is many grow and fast growing species and has potential as a raw
material ‘from its high cellulose content. Cellulose from wood pulp of Sengon
(Paraserianthes faleataria) had be used for making membrane (Rosnelly et al.. 2009,
2010),
Material and Methods.
Materials
Cellulose acetate (CA) used as the membrane forming polymer (acetyl content 39.66% with
molecular weight of 130,221 Da) was produced from erlier researcher by activation,
acetylation, and hydrolysis of wood sengon. The solvent used was dimetyl formamide
61