science and engineering volume - isomase.orgisomase.org/jase/vol.2 may 2015/vol-2.pdf · mohd yazid...

JOURNAL OF AERONAUTICAL

SC I ENCE AND ENG INEER ING

VOLUME . 2

MAY 2 0 , 2 0 1 5

I S S N : 2 4 4 2 6 4 0 7

P U B L I S H E D BY I S OMA s e

Journal of Aeronautical -Science and Engineering-

Vol.2: May, 20 2015

ISOMAse

International Society of Ocean, Mechanical and Aerospace -Scientists and Engineers-

Contents

About JAse

Scope of JAse

Editors

Title and Authors PagesComputational Fluid Dynamic Studies On Helicopter Main-Rotor-Hub Assembly Unsteady Wake

I.S. Ishak, Shuhaimi Mansor,Tholudin Mat Lazim

1 - 8

Computational Fluid Dynamic on Double Delta Wing Khairul Adzha B Abd Halim, Shabudin B Mat

9 – 14

Computational Analysis of a Generic Bell 206B Helicopter Tail Rotor Blade using ANSYS FLUENT

Firdaus, Jaswar Koto, I.S. Ishak, M.S.Ammoo

15 - 22


Vol.2: May 20, 2015

ISOMAse


About JAse

The Journal of Aeronautical -science and engineering- (JAse), ISSN’s registration no: 2442-6407 is an online professional journal which is published by the International Society of Ocean, Mechanical and Aerospace -scientists and engineers- (ISOMAse), Insya Allah, four volumes in a year which are February, May, August and November.

The mission of the JAse is to foster free and extremely rapid scientific communication across the world wide community. The JAse is an original and peer review article that advance the understanding of both science and engineering and its application to the solution of challenges and complex problems in subsea science, engineering and technology.

The JAse is particularly concerned with the demonstration of applied science and innovative engineering solutions to solve specific aeronautic industrial problems. Original contributions providing insight into the use of computational fluid dynamic, heat transfer, thermodynamics, experimental and analytical, application of finite element, aircraft systems, air transportation, air traffic management, and multidisciplinary design optimization of aircraft, flight mechanics, flight and ground testing, flight safety, integration of propulsion and control systems, wing in ground effect, structural design and dynamics, aeroelasticity, and aeroacoustics from the core of the journal contents are encouraged.

Articles preferably should focus on the following aspects: new methods or theory or philosophy innovative practices, critical survey or analysis of a subject or topic, new or latest research findings and critical review or evaluation of new discoveries.

The authors are required to confirm that their paper has not been submitted to any other journal in English or any other language.


Vol.2: May 20, 2015

ISOMAse


Scope of JAse

JAse welcomes manuscript submissions from academicians, scholars, and practitioners for possible publication from all over the world that meets the general criteria of significance and educational excellence. The scope of the journal is as follows:

• Application of Computational fluid dynamic in aeronautic • Heat transfer and thermodynamics • Experimental and analytical of Aircraft • Application of finite element on structural design and dynamics • Aircraft systems, air transportation, air traffic management • Multidisciplinary design optimization of aircraft • Flight mechanics, flight and ground testing, flight safety • Integration of propulsion and control systems • Aeroelasticity and aeroacoustics • General aviation, military and civilian aircraft, WIG, UAV, STOL and V/STOL, subsonic,

supersonic, transonic, and hypersonic aircraft. The International Society of Ocean, Mechanical and Aerospace –science and engineering is inviting you to submit your manuscript(s) to [email protected] for publication. Our objective is to inform the authors of the decision on their manuscript(s) within 2 weeks of submission. Following acceptance, a paper will normally be published in the next online issue.


Vol.2: May 20, 2015

ISOMAse


Editors

Chief-in-Editor

Jaswar Koto (Ocean and Aerospace Research Institute, Indonesia) (Universiti Teknologi Malaysia, Malaysia)

Associate Editors

Adhy Prayitno (Universitas Riau, Indonesia) Ali Selamat (Universiti Teknologi Malaysia, Malaysia) Budhi M. Suyitno (Director General of Air Transportation (ex), Indonesia) Dani Harmanto (University of Derby, UK) Iskandar Shah bin Ishak (Universiti Teknologi Malaysia, Malaysia) Istas Fahrurrazi bin Nusyirwan (Universiti Teknologi Malaysia, Malaysia) Jamasri (Universitas Gadjah Mada, Indonesia) Mazlan Abdul Wahid (Universiti Teknologi Malaysia, Malaysia) Mohd. Shariff bin Ammoo (Universiti Teknologi Malaysia, Malaysia) Mohd Yazid bin Yahya (Universiti Teknologi Malaysia, Malaysia) Musa Mailah (Universiti Teknologi Malaysia, Malaysia) Priyono Sutikno (Institut Teknologi Bandung, Indonesia) Rudy Purwondho (The Institution of Engineers, Indonesia) Sergey Antonenko (Far Eastern Federal University, Russia) Shabudin bin Mat (Universiti Teknologi Malaysia, Malaysia) Tay Cho Jui (National University of Singapore, Singapore) Tresna Priyana Soemardi (Universitas Indonesia, Indonesia) Yasser Mohamed Ahmed Abdel Razak (Universiti Teknologi Malaysia, Malaysia) Wan Khairuddin bin Wan Ali (Universiti Teknologi Malaysia, Malaysia)

Journal of Aeronautical -Science and Engineering-, Vol.2

May 20, 2015

1 Published by International Society of Ocean, Mechanical and Aerospace Scientists and Engineers

Computational Fluid Dynamic Studies On Helicopter Main-Rotor-Hub Assembly Unsteady Wake

I.S. Ishak,a,*, Shuhaimi Mansor, b, Tholudin Mat Lazim,c

Department of Aeronatical, Automotive and Ocean Engineering, Faculty of Mechanical Engineering Universiti Teknologi Malaysia, 81310 UTM Skudai, Johor, Malaysia a)[email protected], b)[email protected], c)[email protected]

*Corresponding author: [email protected] Paper History Received: 10-Apr-2015 Received in revised form: 20-Apr-2015 Accepted: 19-May-2015 ABSTRACT The helicopter tail shake phenomenon is an area of great concern to helicopter manufacturers as it adversely affects the overall performance and handling qualities of the helicopter, and the comfort of its occupants. This study aims to gain information of the flow field that governs helicopter tail shake phenomenon which has frequently puzzled the aerodynamicists. Using Computational Fluid Dynamic (CFD) approach, the Multiple Reference Frames (MRF) method was applied to simulate the helicopter’s main-rotor-hub assembly rotation. In this study, the aerodynamic flow field was computed using the Reynolds-Averaged Navier-Stokes (RANS) equations. As the induced wake, which consequently causing tail to shake differs with the rpm of main-rotor-hub assembly, this preliminary numerical investigation was performed ranging from rpm of 0 to 900, with intervals of 300 rpm. A rotor-hub-fairing which covers partly the main-rotor-hub assembly was also employed to examine its effect on the wake unsteadiness. Results tell the rotation of main-rotor-hub assembly and pylon do significantly influencing the flow unsteadiness. KEY WORDS: Flow Field; Helicopter Tail Shake; Unsteady Flow; Computational Fluid Dynamic. NOMENCLATURE CFD Computational Fluid Dynamic FFT Fast-Fourier Transform GAMBIT Geometry And Mesh Building Intelligent Toolkit

LES Large Eddy Simulation MRF Multiple Reference Frames PSD Power Spectral Density RANS Reynolds-Averaged Navier-Stokes RPM Rotation Per Minute UTM-LST Universiti Teknologi Malaysia - Low Speed Tunnel 1.0 INTRODUCTION Tail shake is an issue of major concern for rotorcraft [1] as it is adversely affected the overall performance and handling qualities of helicopter. Vibrations transmitted from vertical tail to cockpit have also caused discomfort to the occupants [2]. It had being reported on the AH-64D Longbow Apache helicopter, the vibration resulted had increased the cockpit lateral vibration levels, which consequently increased crew workload and reduced their ability to perform precision tasks [3].

This phenomenon is a challenging issue to understand with as it involves an interaction between aerodynamic flow excitation, which related to flight parameters and structural response, which related to structure characteristics [2]. A good understanding of this matter is necessary as a typical aspect of tail shake that it has unsteady random character, indicating that the wake induced excitation is in also unsteady of nature [2].

Tail shake phenomenon happens partly due to the unsteady flow contributed from the main-rotor-hub assembly that hit the tail part, as shown in Figure 1.

Figure 1: Schematic diagram of tail shake [2]


May 20, 2015


Hence, this phenomenon is likely only to happen when there is a forward velocity i.e. helicopter is unlikely to encounter this tail shake problem during hovering or vertical climb/descent,. In this research, focus will be on main-rotor-hub assembly’s wake as it is believed to be the major contributor of the problem. This allows the tail shake investigation to be conducted using blade-stubs configuration [4,5,6]. Blade-stubs configuration is a combination of main-rotor-hub assembly but with shorter rotor blades. 2.0 METHODOLOGY This numerical study used Reynolds-Averaged Navier-Stokes (RANS) equation models which solve ensemble-averaged Navier-Stokes equations [7]. The two-equation models of Realizable k-ε model were employed in this numerical modelling where the Multiple Reference Frames (MRF) approach was applied to simulate the main-rotor-hub-assembly rotation at various rpm.

This numerical investigation used an ellipsoidal fuselage [8] with the axes ratio of longitudinal to lateral axes is 4.485. This kind of model is selected as it avoids geometric complexity and simplifies the interactions with the wake [8]. For this research work, the longitudinal axes were taken as 1120 mm. The model is shown in Figure 2.

Figure 2: An ellipsoidal fuselage [8]

The model is equipped with a simplified main-rotor-hub

assembly attached with shorter blades i.e. blade-stubs configuration, as shown in Figure 3. This simplified main rotor hub assembly is meant to trigger the unsteady wake due to rotation of main- rotor-hub assembly.

Figure 3: A simplified blade-stubs configuration (unit in millimetre)

Figure 4 depicts the complete model with blade-stubs configuration mated with an ellipsoidal pylon.

Figure 4: Model for CFD simulation

In Geometry And Mesh Building Intelligent Toolkit

(GAMBIT), which acts as a pre-processor for CFD analysis, a test section with a size of 2m (width) x 1.5m (height) x 5.8m (length) was virtually created for placing the model. The size of this test section is similar as the test section size of Universiti Teknologi Malaysia – Low Speed Tunnel (UTM-LST), as experimental works are also be planned to compare with this CFD analysis.

In MRF method, two reference frames must be created with one at the vicinity of the rotating parts i.e. main-rotor-hub assembly. As there is no specific rule regarding the size of the frame for rotating parts, this research decided to take the distance of the frame’s boundary to surface of the rotating parts to be 1 aerofoil thickness. The aerofoil thickness of the main rotor, as depicted in Figure 3, is 10 mm and hence the frame was created in such manner as shown in Figure 5.

Figure 5 (a): Frame dimensions (unit in millimetre)

Figure 5(b): Frame for rotating parts

Figure 5: Frame for rotating parts

Ellipsoidal pylon


May 20, 2015


3.0 RESULTS AND DISCUSSION Before running the final simulation, independence grid analysis had been carried out to verify the results obtained are free from grid influence. This is important to confirm the differences of results are due to test configurations, not due to number of the grid.

All simulation results presented throughout this paper were run at wind speed, or forward flight velocity, of 40 m/s and at zero angle of attack.

Figures 6 and 7 show the path lines of turbulent intensity for 0 and 900 rpm of main-rotor-hub assembly, respectively.

Figure 6(a): Top view

Figure 6(b): Side view

Figure 6: Path line of turbulent intensity (%) at 0 rpm

Figure 7(a): Top view

Figure 7(b): Side view Figure 7: Path lines of turbulent intensity (%) at 900 rpm

Viewed from top as shown by Figures 6(a) and 7(a), the path lines are about symmetrical on both left and right sides for 0 rpm, contrary for 900 rpm, the path lines are unsymmetrical indicating there is strong flow interaction with the rotating main-rotor-hub assembly.

Compared to Figure 6(b), turbulent intensity is higher and wake volume is thicker in Figure 7(b). This shows rotation on main-rotor-hub assembly triggers a significant unsteady wake which could contribute to tail shake problem.

Figures 8 and 9 depict the flow unsteadiness at the vicinity of vertical tail i.e. ‘vertical-tail-plane’ for 0 and 900 main-rotor-hub assembly rpm, respectively.

Figure 8: Contours of turbulent intensity (%) at 0 rpm

Figure 9: Contours of turbulent intensity (%) at 900 rpm

Obviously it can be noticed at 900 rpm, there is more flow

unsteadiness compared to at 0 rpm. This justifies the rotor blade rotation does influencing the flow unsteadiness.


May 20, 2015


Table 1 depicts the values of turbulent intensity in this ‘vertical-tail-plane’, ranging from 0 to 900 rpm with intervals of 300 rpm.

Table 1: Turbulent intensity (%)

It indicates main-rotor-hub assembly rpm increases the

turbulent intensity in which is tally with the results of experimental work using a 14% scaled-down generic model of Eurocopter 350Z helicopter done by Ishak et al. [9].

Figure 10 shows total pressure contours for different main-rotor-hub assembly rpm as pressure fluctuations could be translated into aerodynamic forces excitation which may lead to tail vibration.

Figure 10(a): 0 rpm

Figure 10(b): 300 rpm

Figure 10 (c): 600 rpm

Figure 10 (d): 900 rpm

Figure 10: Contours of total pressure (Pa)

The figures depict the fluctuation of total pressure inside the

‘vertical-tail-plane’ is becoming more with the increment of main-rotor-hub assembly rpm. Table 2 summarizes the maximum and minimum values of total pressure inside this ‘vertical-tail-plane’. Table 2: Total pressure (Pa)

Main Rotor Rpm

Total Pressure (Pa) ΔFacet (Pa) Facet max. Facet min.

0 940.24 349.16 591.08 300 940.67 337.17 603.50 600 941.11 324.48 616.63 900 943.08 318.36 624.72

The table indicates the range from minimum to maximum

value of total pressure i.e. Δ Facet, is getting larger with the increment of main-rotor-hub assembly rpm. This result translates more aerodynamic forces excitation occurred with higher rotation of main-rotor-hub assembly, signaling could mean more tail shake problem.

To investigate the effects of covering some parts of main-rotor-hub assembly, pylon height is increased i.e. acting as a rotor-hub-fairing, as shown in Figure 11.

Main Rotor Rpm

Turbulent Intensity (%) Facet max.

0 6.54 300 6.69 600 6.90 900 7.15


May 20, 2015


Figure 11: Model configuration with fairing attached Figures 12 and 13 show path lines with fairing configuration.

Figure 12: Path lines colored by turbulent intensity (%) from

‘main-rotor-hub assembly’ at 0 rpm

Figure 13: Path lines colored by turbulent intensity (%) from

‘main-rotor-hub assembly’ at 900 rpm

The path lines demonstrated by Figures 12 and 13 behave the same trend as with configuration without fairing i.e. more turbulent intensity and wake volume at 900 rpm compares to at 0 rpm configuration.

Compares to Figure 7, Figure 13 translates with fairing configuration, the unsteady wake is reduced both in term of the wake’s unsteadiness and volume.

Figure 14 shows total pressure contours for different rpm of main-rotor-hub assembly with fairing configuration.

Figure 14(a): 0 rpm

Figure 14(b): 300 rpm

Figure 14(c): 600 rpm

Figure 14(d): 900 rpm

Figure 14: Contours of total pressure (Pa)

Same as depicted by Figure 10 for configuration without

fairing, the figures narrate higher pressure fluctuation occurred at higher rotation of main-rotor-hub assembly.

Fuselage Pylon height be increased


May 20, 2015


Table 3 shows the maximum and minimum values of total pressure inside the vicinity of vertical plane with fairing configuration. Table 3:Total pressure (with fairing configuration)

Main rotor rpm

Total Pressure (Pa) ΔFacet (Pa) Facet max. Facet min.

0 936.37 356.93 579.44 300 938.12 351.54 586.58 600 937.99 348.02 589.97 900 937.41 338.21 599.20

Table 3 concludes the same trend as in Table 2 i.e. ΔFacet

increases with the rpm of main-rotor-hub assembly. However its value is smaller compared to Table 2, which interprets lesser aerodynamic excitation happens that might be translated to a lesser tail shake problem.

An investigation was also being carried out trying to correlate the flow unsteadiness with aerodynamic drag, as shown by Figure 15.

Figure 15: Drag on rpm sweep

Figure 15 reveals there is correlation between aerodynamic drag with main-rotor-hub assembly rotation i.e. the aerodynamic drag is increased, even though merely increment, with the incremental of main-rotor-hub assembly rpm. Referred back to Tables 1 and 2 which show main-rotor-hub assembly rpm increases flow unsteadiness, it seems correlation could be made between flow unsteadiness with aerodynamic drag in which higher flow unsteadiness seems generating more aerodynamic drag. This correlation is found agreeable with the results of experimental work done by Ishak et al. [9].

Figure 15 also depicts configuration with fairing is better in term of generating lesser aerodynamic drag. As depicted previously by Figure 14 and Table 3 that fairing leads to lesser flow unsteadiness, this result seems agreeable with the correlation made that flow unsteadiness could influence aerodynamic drag. 4.0 EXPERIMENTAL WORKS In order to verify the simulation findings, experimental works were conducted in the Universiti Teknologi Malaysia's low-speed closed-return wind tunnel (UTM-LST ) as shown in Figure 16. The test section is 2.0m (W) x 1.5m (H) x 5.8m (L) with the

maximum test wind speed of 80 ms-1.

Figure 16: Schematic layout of UTM-LST

4.1 Test Configurations Figure 17 depicts the diagrams for No Fairing Configuration and With Fairing Configuration for the experimental works.

(a) No Fairing Configuration

(b) With Fairing Configuration

Figure 17: Wind tunnel test configurations

It is noted the physical shapes of main-rotor-hub assembly and

the fairing are not exactly as the same as those in the simulation works, but this research is keen more on the results’ trends rather than their absolute values.

The total pressure distribution and its rms value are the main interest of this work as the pressure fluctuations could be translated into wake unsteadiness in the term of aerodynamic forces excitation that may lead to tail vibration.

With the aid of wake rake housing KULITE miniature dynamic pressure transducers type XCL-072-10PSID, the mapping of total pressure was conducted at the downstream of main-rotor-hub assembly. Figure 18 illustrates the schematic diagram of this experimental work and Figure 19 shows the testing be conducted, respectively.

0

2

4

6

8

0 300 600 900

Dra

g (N

)

RPM of Main-rotor-hub assembly

Without Fairing

With Fairing


May 20, 2015


Figure 18: Schematic diagram of the experiment

Figure 19: Experimental works at UTM-LST

However as the scope of this paper is stressed on the simulation

works, only one point of investigation shown in Figure 20 will be analysed for the verification on the findings from simulation works.

(a)No Fairing Configuration

(b) With Fairing Configuration

Figure 20: Schematic diagram for the research point location (the single dotted point)

The sampling rate and number of samples for this experimental works are chosen to be 5kHz and 300 000 data, respectively. This

choice is made after previous study finds these data sampling parameters are appropriate for implementing the dynamic analysis. 4.2 Power Spectral Density (PSD) Analysis Power spectral density (PSD) describes how the power of a signal or time series is distributed with frequency. For this research work, it was done directly by the method called Fast-Fourier Transform (FFT) in which enabling the studies on frequencies and their respective amplitudes.

Analyses done reveal with fairing mated to the main-rotor-hub assembly, there were 69.29% energy reduction of total pressure (Pa2/Hz) and 44.59% reduction on total pressure fluctuations (RMS value). These results coincide and thus verify the numerical findings that conclude fairing does help reducing the wake unsteadiness. 5.0 CONCLUSION This research works serve as a preliminary numerical study on helicopter tail shake phenomenon. Yet using a very simplified model, this research has successfully gives some initial predictions of the flow field that governs the helicopter tail shake phenomenon where the results’ trends are in a good agreement with the results of experimental works done by Ishak et al. [9].

Results tell main-rotor-hub assembly’s rpm influences the flow unsteadiness i.e. at higher rotation, higher turbulent intensity and bigger pressure fluctuations be triggered. This higher aerodynamic excitation likely will contribute to higher helicopter’s tail vibration.

Adding fairing seems will reduce aerodynamic drag, as well the wake’s unsteadiness. This finding could mean a lesser tail shake problem if some of the rotating parts of main-rotor-hub assembly being covered.

Nevertheless, more comprehensive simulation works should be done at various configurations for better comprehensive conclusions. On top of that, Large Eddy Simulation (LES) coupled with Sliding Mesh Method, are much required for a better understanding of this helicopter tail shake phenomenon. ACKNOWLEDGEMENTS Authors would like to thank Abd Basid Abd Rahman, Mohd Riza Abdul Rahman, Airi Ali and Muhammad Khaidhir Jamil, from Universiti Teknologi Malaysia, for their assistance during this research implementation. REFERENCE 1. F.N. Coton (2009). Evaluation Report of UTM Research

Project, University of Glasgow. 2. P.G. de Waad and M. Trouvé (1999). Tail shake vibration,

National Aerospace Laboratory (NLR), American Helicopter Society Annual Forum,.

3. A. Hassan, T. Thompson, E.P.N. Duque, and J. Melton (1999). Resolution of tail buffet phenomena for AH-64DTM Longbow ApacheTM, Journal American Helicopter Society,

Point of Investigation

Wake Pressure Rake

Wind


May 20, 2015


Volume 44. 4. A. Cassier, R. Weneckers and J-M Pouradier (1994).

Aerodynamic development of the tiger helicopter, 50th American Helicopter Society Forum.

5. C. Hermans et al. (1997). The NH90 helicopter development wind tunnel programme, European Aerospace Societies Conference, Cambridge UK.

6. Eurocopter Slide Presentation (2006), France. 7. Introductory FLUENT Notes (2005), FLUENT V6.2. 8. P.F. Lorber, T.A. Egofl (1988). An Unsteady Helicopter

Rotor-Fuselage Interaction Analysis, NASA Contractor Report 4178.

9. I.S. Ishak, S. Mansor, T. Mat Lazim (2008). Experimental Research On Helicopter Tail Shake Phenomenon, Issue 26, Jurnal Mekanikal, Universiti Teknologi Malaysia.


May 20, 2015


Computational Fluid Dynamic on Double Delta Wing

Khairul Adzha B Abd Halim,a, and Shabudin B Mat,a

a)Department of Aeronatics, Automotive and Ocean Engineering, Universiti Teknologi Malaysia, Malaysia Corresponding author:[email protected] Paper History Received: 18-Feb-2015 Received in revised form: 26-Apr-2015 Accepted: 19-May-2015 ABSTRACT

Basic concept of aircraft wing design is based on airfoil section. Time flies, evolution in aircraft wing design shows the desire of mankind to improve the speed, agility, maneuverability of an aircraft. It is proven that aircraft with delta or double delta wing design fulfill this desire. Basic concept of delta wing is triangular planform of the wing that same with the Greek symbol (Δ) and double delta wing is delta wing with a ‘kink’ or leading edge extension. This research aims is to obtain aerodynamic characteristic (Cl, Cd) of double delta wing using computational fluid dynamic and compare the result with wind tunnel experiment. In this work, the geometry of the double delta wing used was constructed using 2 x 106unstructed mesh elements. Turbulence model that been used in this research is k-ω turbulence model. The simulation was run at Reynolds Number of 1 x 106 and 2 x 106 and with variation of pitch angle from 00 to 200. KEY WORDS: Computational Fluid Dynamic; Double Delta Wing. NOMENCLATURE

Cl Coefficient of Lift Cd Coefficient of Drag Re Reynolds Number α Angle of Attack k Turbulent Kinetic Energy ω Specific Dissipation Rate ε Turbulent Dissipation

1.0 INTRODUCTION

The conventional aircraft wing design concept is basically based on airfoil section which comes with low drag at cruising speed. The evolution of aircraft shows that desire of mankind to have a high maneuverability and agility of an aircraft wing design. This desire demands considerable improvements in the aerodynamic characteristics and its related control. To fulfil that desire, many researches had been developed and one of the researches is based on delta wing design. It is best to understand the aerodynamic characteristics from the model it first. Those aerodynamic characteristic can be gain in numbers of ways such as flight test, drop test, water tunnel and even computational fluid dynamic. In this particular study, those aerodynamic characteristic will be obtain by Computational Fluid Dynamic (CFD) simulation. CFD data have potential to be reliable data if a correct turbulent model, geometry, and boundary condition is used. Modelthat will be using in this study is double delta wing with (650/250) sweep angle. 2.0 PAPER FORMAT

2.1 Computational Fluid Dynamic (CFD) Computational Fluid Dynamics or CFD can be described as the use of computers to solve the governing equations for fluid flow in any given situations [1]. CFD represent sets of data for given flow configurations at different Mach number, Reynolds number, etc. same as like wind tunnel. Unlike wind tunnel that heavy, costly, unwieldy device, CFD is much more preferable nowadays because it can be carrying around and can be accessed remotely by people on terminals that can be thousands of miles away from the computer itself.

In order to modelling fluid flow for various geometries, Fluent software is the most preferable software. According to John D Anderson,Fluent supported 2D triangular/ quadrilateral, 3D tetrahedral/ hexahedral/ pyramid meshes and it is also refining or


May 20, 2015


coarsening grid based on the flow solution [2]. There are several turbulent viscous models in Fluent that is appropriate with this thesis such as Spalart-Allmaras model, Standard k-� model and etc. There are few considerations to choice the turbulence model such as the physics encompassed in the flow, level of accuracy required, the available computational resources and the amount of time available for the simulation.

From the several turbulent viscous model, k-ω turbulence model was chosen because from the previous study show that this model can predict the flow separation process with higher accuracy [3]. Wilcox (1998) developed k-ω model to give better compute low Reynolds number effects, compressibility, and shear flow spreading. It is an empirical based model with transport equations for turbulence kinetic energy (k) and specific dissipation rate (ω). Transport equations used in Fluent for Wilcox’s model are as follows.

(1)

(2)

where: Gk - generation of turbulent kinetic energy arises due to mean velocity gradient Gm - generation of rate, w Yk - dissipation of kinetic energy, k Yw - dissipation rate, w αk, αw - turbulent Prandtl numbers 2.1 Double Delta Wing Double delta wing is essentially a delta wing with a “kink” in its leading edges that forms the shoulder where the leading edges of the strake (or Leading Edge Extension, LEX) and main wing intersect [4]. A delta wing is a wing shape when viewed from top like a Greek symbol (Δ) forms like a triangle. It sweeps sharply back from the fuselage with the angle between the leading edge of the wing often high as 60 degrees and the angle between the fuselage and the trailing edge of the wing mostly around 90 degrees.

Aerodynamic investigation of flow over delta wing configurations have been performed for many years. Typical fact that known by previous research are the flow separates already at low angles of attack at the highly swept leading edges [5]. The flow over a delta wing is a vortex dominated flow field. The vortex formed attached to the upper surface of the wing. The flow can be describes as a movement of a part of the flow from the lower to the upper surface into spiral type of motion. This can be seen in Figure 2.1. Verhagen, Jenkins, Kern, and Washburn[6] in their study show that when α < 10o, the two vortices remained separated and hardly interacted. Beyond this angle of attack, the interaction between the two vortices became more pronounced. This was believed to indicate that the breakdown of the strake vortex was causing the wing vortex to burst.

As mention by Lu Zhi-Yong[7] in his study, there are two types of vortex breakdown have been recognized which is in bubble form and the other is the spiral form. Bubble form of breakdown occurs because rapid expansion of the core forming a bubble-like structure that is nearly axisymmetric while for a spiral

form, the vortex centerline deforms into a spiral without any appreciable growth in core size. Figure 2.2 shows the spiral and bubble form of vortex breakdown.

Figure 2.1: Vortex Flow around Double Delta Wing (courtesy from Numerical investigation of high incidence flow over a double-delta wing. Journal of Aircraft, Vol. No 32, 1995).

Figure 2.2: Form of vortex breakdown (courtesy from Study on Forms of Vortex Breakdown over Delta Wing. Chinese Journal of Aeronautics, Vol. 17 No 1, 2004). 3.0 METHODOLOGY

In this chapter, all those sequences during the progress of this thesis will be briefly explained. The idea was divided the whole process into two parts according to the two semesters of study. In this thesis the aerodynamic characteristics obtained by using computer fluid simulation. However, the result from this thesis will be compared with wind tunnel experiment results.


May 20, 2015


3.1 Flow Chart

Figure 3.1: Process of CFD simulation

3.2 Pre-Processing Pre-processing is where to prepare the input data before a simulation is run. At first, by using others third party software like SolidWorks to set up the geometry. The solid drawing of this particular double delta wing was given in Figure 3.2. Then, this model was subtracted from a block of 5800 mm x 2000 mm x 1500 mm which is similar to the control volume in testing section of UTM Low Speed Tunnel. The solid area will be representing as the fluid moving around the model.Block with cavity of the double delta wing then will be cut to be half since we only consider pitch angle in this experiment. There is not consideration of yaw angle in this experiment. The cavity will be half body of double delta wing.

Figure 3.2: Double Delta Wing Model

Figure 3.3: Half Body of Double Delta Wing

The solid cavity block then meshed to direct the flow around

the model. The solid must be converted first as a paraSolid before being import to ANSYS Workbench. The best grid was chosen in order to get the best result. Too much fine grid will contribute to a longer time as the computer need more time to compute the flow field.

3.3 Post Processing This is where value of boundary condition was defined and iteration calculation started. In this study, FLUENT will be used as the solver and 3d type of analysis was chosen.

Table 3.1: Boundary Condition Re 1x106 2x106

(αo) x-component (m/s)

y-component

(m/s)

x-component

(m/s)

y-component

(m/s) 0 30.51 0 61.02 0 5 30.394 2.659 60.788 5.318

10 30.046 5.298 60.093 10.596 15 29.470 7.897 58.941 15.793 20 28.670 10.435 57.340 20.870

3.4 Result Simulation All the results can be show either by contour of pressure, density, velocity and many more. The results also can be save in txt file to do further analysis and comparison for other result. CFD simulation and wind tunnel measurement were compared associated with the difference of all forces and the moment. All data will be tabulated and graph of data analysis between CFD simulation and Wind Tunnel test will be plot. 4.0 RESULT

4.1 Introduction This part shows the result that had been gained in from the computational fluid dynamics. The main concern for this study is to get the flow visualization and determined the aerodynamic characteristics of the double delta wing. Comparison between CFD result with wind tunnel testing focused on aerodynamic characteristic which is lift coefficient and drag coefficient.


May 20, 2015


4.2 Flow Visualisation Figure 4.1 shows the velocity contour of double delta wing at 20% of chord. Range for contour velocity is made to be fixed from 0 to 42 m/s to make comparison from various angles of attack. As we can see from the contour of velocity of the model, it is shown that as the angle of attack increases, the vortex formed by the model is become more severe.

Figure 4.2 show the contour of total velocity of double delta wing at 70% of chord while Figure 4.3 shows the contour of total pressure of double delta wing at 70% of chord. Basic of fundamental fluid dynamic that say when velocity is higher at certain place, the pressure at the place will be the lowest and vice versa. This phenomenon can be seen from the Figure 4.2 and Figure 4.3.

Figure 4.1: Front view contour of total velocity at 20% chord at Re = 1 x 106

Figure 4.2: Front view contour of total velocity at 70% chord at Re = 1 x 106

Figure 4.3: Front view contour of total pressure at 70% chord at Re = 1 x 106


May 20, 2015


4.3 Aerodynamic Characteristics As mention before, aerodynamic characteristic is the most important data for aircraft flight. Lift and drag coefficient will be one of determination of the performance of an aircraft. Equation 4.1 and equation 4.2 was used to gain those aerodynamic characteristics.

(3)

(4)

Table 4.2: Lift and Drag Coefficient of double delta wing using CFD Reynolds Number

1 x 106 2 x 106

Angle of attack (deg)

CL CD CL CD

0 0.006668 0.065309 0.012645 0.044611 5 0.199926 0.107474 0.141193 0.098939 10 0.504066 0.139798 0.414291 0.151727 15 0.680203 0.212802 0.701801 0.196719 20 0.901917 0.41173 1.1046 0.380174

Figure 4.4 shows behaviors of lift coefficient react with angle

of attack at different Reynolds number. Lift curve slop develop from 2 x 106 Reynolds number seems more higher compared to 1 x 106 Reynolds number. Figure 4.5 shows variation of drag coefficient at different angle of attack at one and two million of Reynolds number. This shows that Reynolds number did not give so much effect on drag coefficient.

Figure 4.4: Lift coefficient at different Reynolds Number

Figure 4.5: Drag coefficient at different Reynolds Number

4.4 Comparison Result with Wind Tunnel Testing In wind tunnel experiments, six-component external balance is used to get the value of the aerodynamic loads such as lift, drag, side force, pitching moment, rolling moment and yawing moment. In this particular study, we only consider the lift and drag force since this two aerodynamic characteristic is our main concern. Table 4.3: Comparison of lift coefficient (CL) between CFD and Experimental


CFD Result EXP Result ΔCL

0 0.0066 -0.0641 1.103 5 0.1999 0.3485 0.426

10 0.5040 0.7188 0.298 15 0.6802 0.9073 0.250 20 0.9019 1.0781 0.613

Table 4.4: Comparison of drag coefficient between CFD and Experimental


CFD Result EXP Result ΔCD

0 0.0066 -0.0641 0.1876 5 0.1999 0.3485 0.2198

10 0.5040 0.7188 0.2688 15 0.6802 0.9073 0.0489 20 0.9019 1.0781 0.1401

Figure 4.6: Comparison lift coefficient at Re = 1 x 106

Figure 4.7: Comparison drag coefficient at Re = 1 x 106


May 20, 2015


Based on both figure for lift and drag coefficient for CFD result and wind tunnel result shows similar pattern and trends of graph. In addition, Table 4.2 and Table 4.3 show the value of error gain from CFD result compare to the wind tunnel testing. The highest percentage error in lift coefficient is at 00 angle of attack with almost 100 percent error. This is due to CFD result gain positive value while experimental gain negative value. However, well known that at 00 angle of attack, the lift coefficient is supposed to be zero and both of CFD and experimental result give value near to zero. 5.0 CONCLUSION In this final year project, an aerodynamic study of double delta wing was performed by using Computational Fluid Dynamic (CFD) code Fluent. This study was focusing on aerodynamic characteristic of the model. Scope of this study is about literature review on CFD and double delta wing, simulation of CFD on double delta wing and comparison of the result from CFD with wind tunnel experiment.

A simulation by CFD has been carried out on a 65/25 degree of double delta wing at angles of attack ranging from 0 to 20 degree and at Reynolds number for one million and two million. A grid independent study was carried out to get an accurate result. It is shown that mesh model with over one million of element will be give better result. A mesh model can be varied from 200 thousand to 2 million of no of elements.

The error may happen due to less mesh quality in simulation process. Even a slight change in mesh quality will affect the simulation result. There are two types of mesh in CFD software that is structured mesh and unstructured mesh. The mesh model use in this simulation is unstructured mesh. This is because unstructured mesh consumes less time and much easier to handle compared to structured mesh. Another factor affecting the result between the CFD and experimental may come from human error while conducting the experiment.

It is shown that double delta wing can achieve high maneuverability and agility in its performance. Therefore, for further study, firstly, additional angle of attack for this model to see the characteristic since this study only simulate until 200 of angle of attack. From there, more result will be obtained for the agility of the wing itself. Secondly, highly recommended one can simply make a structured mesh of double delta wing model in order to gain higher quality of mesh model. Besides that, it is recommended to run the simulation with difference turbulence model such k-ε, spalart-allmaras, etc.

Then, for further studies of this double delta wing model, some configuration can be made such as instead of using sharp leading edge, a blunt leading edge can be used. Other than that, adding a control surface to the model such as flap to see how it will affect the aerodynamic characteristic of the model. Furthermore, it is also recommended to make a model of double delta wing with difference sweep angle combination such as 650/450 or 700/400 to compare and study the differences.

ACKNOWLEDGEMENTS The authors would like to convey a great appreciation to Department of Aeronatical, Automotive and Ocean Engineering, Universiti Teknologi Malaysia, Malaysia for supporting this research. REFERENCE 1. Shaw, C. T. (1992). Using Computational Fluid Dynamics.

Prentice Hall. 2. John D. Anderson, J. (2011). Fundamental of Aerodynamics.

New York, McGraw Hill, Inc. 3. S. Saha, M. (2012). Flow Visualization and CFD Simulation

on 650 Delta Wing at Subsonic Condition. Jadavpur University, Kolkata, India.

4. Nettelbeck, C. (2008). Dynamic Analysis of a Double Delta Wing in Free Roll. School of Aerospace, Civil and Mechanical Engineering, (BE)

5. Breitsamter, A. F. C. (2012). "Turbulent and Unsteady Flow Characteristics of Delta Wing Vortex Systems."

6. Verhaagen N. G., Jenkins L. N., Kern S. B., and Washburn A. E. A study of the vortex flow over a 76/40-deg double-delta wing. AIAA-1992-279 33rd Aerospace Sciences Meeting and Exhibit, Reno NV, 1995.

7. Lu Zhiyong, Zhu Lirguo (February 2004), Study on Forms of Vortex Breakdown over Delta Wing

8. Luckring, J. M. (2010). A Survey of Factors Affecting Blunt Leading-Edge Separation for Swept and Semi-Slender Wings. AIAA 28th Applied Aerodynamics Conference. Chicago

9. John D. Anderson, J. (1995). Computational Fluid Dynamic, The Basics With Applications. United State of America, McGraw-Hill, Inc

10. Hesamodin Ebnodin Hamidi, M. R. (2011). Numerical Investigation of High Attach Angle Flow on 76o/45o Double Delta Wing in Incompressible Flow. World Academic of Science, Engineering and Technology.

11. N. Verma, D. S. S. (September 2012). "Spalart Allmaras Unsteady Flow Investigation Using Computational Fluid Dynamics." International Journal of Engineering Research & Technology (IJERT) Volume 1(Issue 7).


May 20, 2015


Computational Analysis of a Generic Bell 206B Helicopter Tail Rotor Blade using ANSYS FLUENT

Firdausa, Jaswar Kotoa,b,*, I.S. Ishak a and M.S. Ammooa

a)Department of Aeronautical, Automotive and Ocean Engineering, Faculty of Mechanical Engineering, Universiti Teknologi Malaysia b)Ocean and Aerospace Engineering Research Institute, Indonesia *Corresponding author: [email protected] and [email protected]

Paper History Received: 15-May-2015 Received in revised form: 18-May-2015 Accepted: 19-May-2015

ABSTRACT This paper presents the computational analysis of three dimensional (3D) flow over the generic model of Bell 206B helicopter tail rotor blades using ANSYS Fluent Computational Fluid Dynamics (CFD) package. This simulation work deals with the comparative study of variation in an angle of attack over the blade at different speed which using the k–ω shear stress transport (SST) model. The model is utilized to predict the flow accurately along with turbulence intensities 5% and 5% at velocity inlet and pressure outlet respectively. The meshing 3D geometry was performed on ICEM CFD package of ANSYS and the simulation was executed in the FLUENT package of ANSYS. The simulation results were validated by comparing with the experimental result that have already done before.

KEY WORDS: NACA 0012; lift coefficient (CL); drag coefficient (CD); helicopter tail rotor blades; Bell B206; k–ω shear stress transport (SST) model; FLUENT.

NOMENCLATURE CL Lift coefficient CD Drag coefficient CFD Computational Fluid Dynamic DES Detached Eddy Simulation SST Shear Stress Transport

1.0 INTRODUCTION The propeller blade is the device that mainly used as propulsive for marine vehicles, airplanes and rotorcraft. As it is a crucial part, it has to be designed to meet power requirement at the indicated speed with optimum efficiency. Now days, with growing demands for of higher speed and greater power, the propeller is becoming increasingly larger in size and its geometry shape become more complicated. Due this complicated geometry, the propeller should be optimally designed for increased propulsion efficiency.

To predict the steady and unsteady propeller characteristics, many numerical models for propeller blade simulation were proposed.

Recently, advanced Reynolds Averaged Navier Stokes (RANS) equations are well known for numerically predicting fully turbulent part of the flow field. Even though much accomplishment has been achieved, essential problem still exists where the standard fully turbulent RANS approach fail to give sufficiently accurate results (Niels N. Srensen, 2009). This problem is due to, none of these models are sufficient to handle flows with significant transition effects because of lack of practical transition modeling (Ünver Kaynak, 2012). Nevertheless, transition predictions have shown certain progress and utility by means of the well-known eN method (Giles, 1987), some two-equation low Re-number turbulence models (Wilcox, 1994), and some methods based on experimental correlations (Suzen, 2000).

The new correlation based γ−Re� model (Menter et al., 2004) and the Detached Eddy Simulation (DES) version of the k − ω SST model (Strelets, 2001) is applied, in order to execute the flow simulations. It is well known that the movement of the separation point on the airfoil is highly influenced by the laminar to turbulent transition process. Furthermore, it is well known that typical RANS are not sufficiently accurate in massively separated flows, and the DES technique is applied to solve this problem.

In this study, 3-D computational result was obtained using the FLUENT software for a generic model of Bell 206B helicopter tail


May 20, 2015


rotor blade airfoil. The aim of this study is to investigate the prediction of aerodynamic characteristics of the tail rotor blades. The k-ω shear stress transport (SST) transition model are used in conjunction with the build-in Reynolds-averaged Navier-Stokes (RANS) solver. The results were compared with the results gathered from previous experiment of a full-scaled generic model of Bell 206B helicopter tail rotor blade, conducted in the Universiti Teknologi Malaysia- Low Speed Tunnel (Firdaus, 2015). 2.0 THEORETICAL FORMULATIONS NUMERICAL METHOD 2.1 The k-ω Shear Stress Transport (SST) Model This model was implemented from two-equation Eddy-Viscosity turbulence models that were developed by Menter, F. R (1994) to efficiently blend the vigorous and precise formulation of the k-ω standard model in the near-wall region with the free-stream liberation of the k-ω standard model in the far field. This is achieved by the conversion of the k-ω standard model into a k-ω formulation (Tousif Ahmed, 2013). The k-ω SST model is comparable to the standard k-ω model, but following improvement are included: a) Both the standard k-ω model and the transformed k-ω model

was multiplied by a blending function and both models are added together after that. The blending function is the one activating the standard k-ω model in the near-wall region, and it is zero away from the surface, which activates the transformed k-ω model.

b) The SST model integrated a damped cross-diffusion derivative term in the ω equation.

c) The transport of the turbulent shear stress is accounted from the turbulent viscosity definition that has been modified.

d) The modeling constants are made different. These features make the SST k-ω model more accurate and

reliable for a wider class of flows (e.g., adverse pressure gradient flows, airfoils, and transonic shock waves) compared to the standard k-ω model. SST k-ω model is given by the following:

�(��)

� + �(��)� � = � − �∗�� + �

� � �(� + ��) �� (1)

�(��)

� + �(��)� � = �

�� − �� + �� (� + ��) ��

� �� + 2(1 − "#) �$%&

��

��

(2) Where

� = '() *�(*+)

'() = � ,2-() − 23 *��*+� /()0 − 23 ��/()

-() = 12 1*�(*+) + *�)*+( 2

and the turbulent eddy viscosity can be shown as:

� = �3#�max (3#�, Ω"�

(3) An inner (1) and outer (2) constant was blended for each of the constants are done by: 9 = "#9# + (1 − "#)9� (4) Where 9# is constant 1 and 9� is constant 2. And the other functions are given by: "# = tanh(=>?#@)

=>?# = min BC=+ 1 √��∗�E , 500HE�� 2 , 4��JK��E�L JK�� = C=+ 12�� 1� *�*+)

*�*+) , 10M�N2

"� = tanh(=>?��) =>?� = max B2 √��∗�E , 500HE�� L Where � is the density, H = � /� is the turbulent kinematic viscosity, � is the molecular dynamic viscosity, d is the distance

from the field point to the nearest wall, and P = Q2R()R() is the

vortices magnitude, where

R() = 12 1*S(*+) − *S)*+( 2

(5) Lastly, the model constant are: ��# = 0.85 3�# = 0.5 �# = 0.075 �� = 1.0 3�� = 0.856 �� = 0.828 �∗ = 0.09 � = 0.41 3# = 0.31 2.2 Transition Model The k–ω SST transition model (Menter, F.R, 2004) as implemented within the RANS equations; solves for four transport equations, such as the turbulent kinetic energy (k), specific turbulence dissipation rate (ω), intermittency (γ), and the transition onset momentum thickness Reynolds number (Re�T) equations.

The correlation between γ transport equation and Re�U transport equation are based transition model developed by Menter (2004). This framework is for the implementing empirical correlations based transition criteria in general purpose flow solvers, where, structured, unstructured and parallelized solvers can be used together (Niels N. Srensen, 2009). The constancy of this model is two transport equations, one for intermittent γ and one for the local

transition onset momentum thickness Reynolds numberReU�T . Mostly, the model relates the local momentum thickness Reynolds number Re� to the critical valueRe�V , and switches on the intermittency production when the Re� is larger than the local critical value. Based on series of zero pressure gradient flat plate boundary layers, Sørensen (2008) have been


May 20, 2015


determined the expressions for two missing correlation functions

relating Re�V and "WXYZ[to ReU �T. The equation dependency of the

two correlations is approximated by the following expression:

\]^_ = � 1\]^̀ + 1200025 2 + (1 + �) 17. \]^̀ + 10010 2

(6)

"WXYZ[ = min c150. exp c− 1\]^̀120 2#.�e + 0.1,30e

(7) Where � is a blending function defined as:

� = tanh c1\]^̀ − 100400 2@e (8)

Figure 1 shown good agreement comparing between the present correlation for Re�V with the two correlations proposed by Toyoda et al (2007) and Pettersson et al. (2008) at low Re�T. Toyoda et al. stated in his paper that the expression for the "WXYZ[ parameter has

dimension of length, and not a dimensionless quantity as it should be. A direct comparison of the expression of Toyoda et al. and the present "WXYZ[parameter therefore is not possible, other way round

the correlation proposed by Pettersson et al. show good agreement, shown in Figure 1.

Figure 1: Comparison between the four target points from the optimization and different correlation functions.

Note that, in the case of zero shear where there is no production in the far field, as go to the turbulent will decay from the inlet value. It is shown in Eqn. 10, the farfield value can be estimated to control the level of turbulent kinetic energy at the boundary layer edge, reference (R.B Langtry, 2006):

� = �(YWX(1 + �(YWX�f)Mg∗g ,

(9) where � = 0.09, and �∗ = 0.828 3.0 AXIAL, NORMAL, LIFT, AND DRAG FORCE DIRECTIONS PROCEDURE The force coefficient FX and FY are parallel and perpendicular to the chord line of the blade, whereas the more usual coefficient FL and FD are defined with reference to the direction of the free-stream airflow. (E.L Houghton, 2013)

Figure 2: Definition: axial, normal, lift, and drag force directions.

The conversion from one pair of coefficient to the other may be carried out with reference to Figure 7 which is FR, the coefficient of the resultant aerodynamic force, act at an angle γ to FY. FR is the result both of FX and FY and of FL and FD: therefore, based on the Figure 2, it can defined that FL = FRcos (γ + α) = FRcos γ cos α – FR sin γ sin α (10) But FRcos γ = FY and FR sin γ = FX, so

The lift force can be calculated by:

FL = FYcos α – FX sin α (11)

Similarly, the drag force also can be calculated by:

FD = FR sin (γ + α) = FY sin α + FXcos α (12)

And finally, the coefficients are given by the relationships

Lift coefficient, Jn = "n#� �o�-

(13)


May 20, 2015


Dragcoefficient, Js �"s

#

��o�-

(14)

4.0 DESCRIPTION OF THE PHYSICAL MODEL 4.1 Airfoil Blade The model was based on the actual scale of a generic model of the tail rotor blades Bell 206B helicopter, as shown in Figure 3. The blade was 720 mm overall length and has 134 mm length of the chord.

The airfoil profile of the blade is near similar to NACA 0012 series, with maximum thickness 12% at 33% chord as shown in Figure 4. In this study, 3D CAD geometry blade model as shows in Figure 5 has been generated in AutoCAD will be used in this simulation.

Figure 3: The generic used model of tail rotor blade Bell B206

Figure 4: Airfoil profile of tail rotor blade. 4.2 Solution Grid The 3D CAD model of the tail rotor blade was imported in the Design Modeler and the 3D block domain has been generated, as shown in Figure 6. Figure 7 and Figure 8 shows the meshing of 3D block structured hexahedral grids were created in the pre -processor ICEM-CFD. This pre-processor is the computer program were can be used to generate structured or unstructured meshes consisting of quadrilateral, triangular or tetrahedral elements of 2D and 3D models. In this case, the mesh for the blade model is an unstructured type, consisting 2,536,280 cells and 630,727 nodes.

Figure 5: The 3D CAD model of tail rotor blade from Bell B206

using AutoCAD.

Figure 6: 3D block geometry in ANSYS Design Modeler

Figure 7: 3D block structured hexahedral mesh


May 20, 2015


Figure 8: Close-up view of the meshing around the blade model 5.0 SIMULATION METHOD In order to validate the present simulation process, the operating conditions are mimicked to match the operating conditions of the experimental works conducted previously in the Universiti Teknologi Malaysia- Low Speed Tunnel (Firdaus, 2015). The velocity inlet (Figure 9) for the simulation were set from 5 m/s to 40 m/s, corresponds to a Reynolds number based on airfoil chord from 0.419 × 105 to 3.352 × 105 and angle of attack of 0, 5, 10, 12, 15, 18, 20, and 25 degrees. The free stream temperature is 298.65 K, which is the same as the surrounding air temperature. The density of the air at the given temperature is ρ = 1.17kg/m3 and the viscosity is μ = 1.859 × 10Mwkg/m-s.

In this study, it is assumed that inlet velocity is same turbulent as pressure outlet. So, for velocity inlet boundary condition turbulent intensity is considered 5% and for pressure outlet boundary is also considered 5% as recommended by ANSYS, which state the inlet boundary condition turbulence intensities is ranging from 1% to 5%. In addition, turbulent viscosity was set to 10 for a better approximation of the problem as recommended by ANSYS.

Table 1 shows the ANSYS FLUENT setup for k–ω SST transition model before the simulation are executed:

Figure 9: Domain mesh of FLUENT simulation

SOLUTION SETUP

GENERAL Solver Type Pressure-Based

Velocity Formulation

Absolute

Time Steady

MODELS Viscous Transition SST

MATERIALS Air Density = 1.17 kg/m3

Viscosity = 1.859 × 10-5 kg/m-s

BOUNDARY CONDITIONS Gauge Pressure 0 Pascal

Temperature 298.65 K

Operating Pressure 101325 Pascal

Turbulence Specification

Method Intermittency, Intensity and Viscosity Ratio

Intermittency 1

Turbulent Intensity 5 % Turbulent

Viscosity Ratio 10

SOLUTION

SOLUTION METHODS Pressure-Velocity Coupling

Scheme Coupled

Spatial Discretization Gradient Least Square Cell Based Pressure Second Order

Momentum Second Order Upwind Turbulent Kinetic

Energy Second Order Upwind

Momentum Thickness Re

Second Order Upwind

SOLUTION CONTROLS Explicit Relaxation Factors

Momentum 0.75 Pressure 0.75 Density 1

Body Force 1 Turbulent Kinetic

Energy 0.8

Momentum Thickness Re

0.8

Turbulent Viscosity

1

Table 1: ANSYS FLUENT setups for k–ω SST transition model


May 20, 2015


6.0 RESULT AND DISCUSSION

Drag coefficient Lift Coefficient

Figure 10 (a): Drag and lift coefficient of the blade model at different air speed and angles of attack.

0.0

0.2

0.4

0.6

0 5 10 15 20 25

CO

EF

FIC

IEN

T O

F D

RA

G(C

D)

Angle of Attack

CD at Velocity = 15 m/s

15 m/s-Experiment 15 m/s-ANSYS (FLUENT)

-0.5

0.0

0.5

1.0

1.5

0 5 10 15 20 25

CO

EF

FIC

IEN

T O

F L

IFT

(CL)

Angle of Attack

CL at Velocity = 15 m/s


0.0

0.2

0.4

0.6

0 5 10 15 20 25

CO

EF

FIC

IEN

T O

F D

RA

G(C

D)

Angle of Attack



-0.5

0.0

0.5

1.0

1.5

0 5 10 15 20 25

CO

EF

FIC

IEN

T O

F L

IFT

(CL)

Angle of Attack



0.0

0.2

0.4

0.6

0 5 10 15 20 25

CO

EF

FIC

IEN

T O

F D

RA

G(C

D)

Angle of Attack



-0.5

0.0

0.5

1.0

1.5

0 5 10 15 20 25

CO

EF

FIC

IEN

T O

F L

IFT

(CL)

Angle of Attack




May 20, 2015


Drag coefficient Lift Coefficient

Figure 10 (b): Drag and lift coefficient of the blade model at different air speed and angles of attack. Figure 10 (a) and (b) show the comparison of CD and CL values between numerical and experimental works for variation of wind speed and angle of attack.

It is observed that the numerical results give reasonably good tendency against the ones from experimental works. This tendency starts improving with an increment of Reynolds number. The

results for the drag coefficient (CD) show that there is quite good agreement between the experiment and computational results. On the contrary, the prediction of computation for the lift coefficient (CL) slightly overshoot against the experimental results.

Although the predictions of k–ω SST transition model do not fully agree, the relative agreement is still reasonable as both

-0.5

0.0

0.5

1.0

1.5

0 5 10 15 20 25

CO

EF

FIC

IEN

T O

F L

IFT

(CL)

Angle of Attack



-0.5

0.0

0.5

1.0

1.5

0 5 10 15 20 25

CO

EF

FIC

IEN

T O

F L

IFT

(CL)

Angle of Attack



0.0

0.2

0.4

0.6

0 5 10 15 20 25

CO

EF

FIC

IEN

T O

F D

RA

G(C

D)

Angle of Attack



0.0

0.2

0.4

0.6

0 5 10 15 20 25

CO

EF

FIC

IEN

T O

F D

RA

G(C

D)

Angle of Attack



-0.5

0.0

0.5

1.0

1.5

0 5 10 15 20 25

CO

EF

FIC

IEN

T O

F L

IFT

(CL)

Angle of Attack



0.0

0.2

0.4

0.6

0 5 10 15 20 25

CO

EF

FIC

IEN

T O

F D

RA

G(C

D)

Angle of Attack




May 20, 2015


computational and experimental, at least agrees at the trend of the lift and drag coefficients as the Reynolds number changes.

In case for low Reynolds number flow conditions, predicting the drag coefficient becomes even more problematic, since low Reynolds number airfoils normally exhibit laminar separation bubbles, which are known to significantly affect the performance of an airfoil blade. As for increasing the Reynolds number, the ability of prediction of drag and lift coefficient becomes easier for the flow solvers because the flow is no longer laminar, and turbulent boundary layer is effective on the surface of the airfoil blade beginning from the leading edge. 7.0 CONCLUSION The computational analysis of three dimensional (3D) flow over the generic model of Bell 206B helicopter tail rotor blades using ANSYS FLUENT based on k–ω SST transition model has been demonstrated. The solutions obtained from FLUENT simulation are compared with the experimental result (Firdaus, 2015), and it is noted that the prediction gives a good agreement for the inclination of the lift coefficient and drag coefficient, although over predict at low Reynolds number flow conditions. Nevertheless it is conceded that there could be discrepancies with the exact data of the 206B rotor tail blade since several assumptions had been made, and also due to some limitations of experimental and simulation works. REFERENCE 1. Niels N. Srensen, 2009, “3D CFD Computations of

Transitional Flows Using DES and A Correlation Based Transition Model”. Risø DTU. National Laboratory for Sustainable Energy. Technical University of Denmark

2. Ünver Kaynak Samet Çaka Çakmakçıoğlu1 and Mustafa SerdarGenç, 2012, “Transition at Low-Re Numbers for some Airfoils at High Subsonic Mach Numbers”. Low Reynolds Number Aerodynamics and Transition, InTech Europe, pp. 79-96.

3. Giles, Michael B. and Mark Drela, 1987, “Two-Dimensional Transonic Aerodynamic Design Method”,AIAAJournal,vol. 25, no. 9, Sept. 1987, pp. 1199–1206.

4. Wilcox, D.C., 1994, “Simulation of Transition with a Two-Equation Turbulence Model”, AIAA Journal, Vol. 32, No. 2, pp. 247-255.

5. Suzen, Y. B. and Huang, P.G., 2000, “Modeling of Flow Transition Using an Intermittency Transport Equation”, ASME Journal of Fluid Engineering, Vol. 122, pp. 273-284.

6. F. R. Menter, R. B. Langtry, S. R. Likki, Y. B. Suzen, P. G. Huang, and S. Volker, 2004, “A Correlation-Based Transition Model Using Local Variables, Part I - Model Formulation”. In Proceedings of ASME Turbo Expo 2004, Power for Land, Sea, and Air, Vienna, Austria, June 14-17

2004. ASME. GT2004-53452. 7. M. Strelets, 2001, “Detached Eddy Simulation of Massively

Separated Flows”. AIAA Paper 2001-0879, Russian Scientific Center”Applied Chemistry” St. Petersburg.

8. Firdaus, Jaswar Koto, I.S Ishak, M.S Ammoo, 2015, “Wind Tunnel Test on Generic Agusta-Bell 206B Helicopter Tail Rotor Blades”. Journal of Aeronautical -Science and Engineering-, Vol 1, pp. 1-6.

9. Tousif Ahmed , Md. Tanjin Amin , S.M. Rafiul Islam &Shabbir Ahmed, 2013, “Computational Study of Flow Around a NACA0012 Wing Flapped at Different Flap Angles with Varying Mach Numbers”. Global Journal of Researches in Engineering General Engineering. Volume 13 Issue 4 Version 1.0.

10. Menter, F.R., Langtry, R.B., Likki, S.R., Suzen, Y.B., Huang, P.G. and Völker, s. , 2004, “A Correlation Based Transition Model Using Local Variables: Part I-Model Formulation ASME-GT2004-53452, Proceedings of ASME Turbo Expo 2004, Vienna, Austria, pp. 57-67.

11. E.L. Houghton, P.W. Carpenter, Steven Collicott and Dan Valentine. 2012. Aerodynamics for Engineering Students (Sixth Edition). Butterworth-Heinemann. Page: 54-55.

12. A. Toyoda, T. Misaka, and S. Obayashi, 2007, “An Application of Local Correlation-Based Transition Model to JAXA High-Lift Configuration Model”. AIAA Paper 2007-4286, June 2007.

13. K. Pettersson and S. Crippa, 2008, “Implementation and Verification of a Correlation Based Transition Prediction Method”. AIAA Paper 2008-4401.

14. R. B. Langtry, J. Gola, and F. R. Menter,2006, “Predicting 2D Airfoil and 3D Wind Turbine Rotor Performance using a Transition Model for General CFD Codes”. AIAA-paper-2006-0395.

Pub

ISOMRestyJalan Pekanhttp://

Pub

TekniFakulUnivehttp://

MechInstituINDO

blished

MAse y Menara Hotel

Sisingamangarnbaru-Riau, IND/www.isomase

blication

ik Mesin ltas Teknik, ersitas Riau, IN/www.unri.ac.i

hanical Chapter ution of Engine

ONESIA

raja No.89 DONESIA .org/

NDONESIA d/en

of the eers,

Suppor

Ed

BuDeAuFaTe ht

O

OcInsPekhtt

rted

MaInsInsSc-So&I

dited

uilding P-23, Roepartment of Aeutomotive & Occulty of Mecha

eknologi Malaysttp://web1.fkm.u

rganized

cean & Aerospastitute, Indonesikanbaru-Riau, Itp://isomase.org

alaysian Joint Bstitution of Navstitute of Marinience and Techouthern ChapteIMarEST – SC)

oom: 314 eronautics, cean Engineerinanical, Universitsia, MALAYSIutm.my/

ace Research ia INDONESIA

g/OCARI.php/

Branch Royal val Architects &ne Engineering, hnology er (MJB RINA )-

ng, ti IA

&

ISSN: 24422-6407

science and engineering volume - isomase.orgisomase.org/jase/vol.2 may 2015/vol-2.pdf · mohd yazid...

Documents