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 mandirEDITORIAL 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 STNOTICE 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, IndonesiaMESSAGE 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 UniversityMESSAGE 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 InstituteCONTENTS 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 99 1127 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 122 127 134 140 145 153 160 165 174 184 190 195 202 210 218Proceeding 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, 53Proceeding 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). 54Proceeding 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). 55Proceeding 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@_—— 56Proceeding 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 57Proceeding 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 58Proceeding 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 59Proceeding 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, 60Proceeding 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