vol.3, january.2014 issn 2354-7065isomase.org/jomase/vol.3 jan 2014/vol-3.pdf · buana ma’ruf...

ISSN 2354-7065Vol.3, January.2014

Journal of Ocean, Mechanical and Aerospace -Science and Engineering-

Vol.3: January 2014

ISOMAse

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

Contents

About JOMAse

Scope of JOMAse

Editors

Title and Authors PagesCrack Growth Simulation at Welded Part of LNG Tank

Kazuhiro Suga, Takafumi Endo and Masanori Kikuchi

1 - 7

Study of the Effect of Low Profile Vortex Generators on Ship Viscous Resistance

Yasser M. Ahmed, A. H. Elbatran and H. M. Shabara

8 - 14

Design of a Body with Depth Control System for an Underwater Glider Muhamad Fadli Ghani and Shahrum Shah Abdullah

15 - 17

Performance of VLCC Ship with Podded Propulsion System and Rudder Jaswar Koto and Amirul Amin

18 - 24


Vol.3: January 2014

ISOMAse


About JOMAse

The Journal of Ocean, Mechanical and Aerospace -science and engineering- (JOMAse, ISSN: 2354-7065) is an online professional journal which is published by the International Society of Ocean, Mechanical and Aerospace -scientists and engineers- (ISOMAse), Insya Allah, twelve volumes in a year. The mission of the JOMAse is to foster free and extremely rapid scientific communication across the world wide community. The JOMAse 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 naval architecture, offshore and subsea, machines and control system, aeronautics, satellite and aerospace. The JOMAse is particularly concerned with the demonstration of applied science and innovative engineering solutions to solve specific industrial problems. Original contributions providing insight into the use of computational fluid dynamic, heat transfer, thermodynamics, experimental and analytical, application of finite element, structural and impact mechanics, stress and strain localization and globalization, metal forming, behaviour and application of advanced materials in ocean and aerospace engineering, robotics and control, tribology, materials processing and corrosion generally 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.3: January 2014

ISOMAse


Scope of JOMAse

The JOMAse 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:

• Environment and Safety • Renewable Energy • Naval Architecture and Offshore Engineering • Computational and Experimental Mechanics • Hydrodynamic and Aerodynamics • Noise and Vibration • Aeronautics and Satellite • Engineering Materials and Corrosion • Fluids Mechanics Engineering • Stress and Structural Modeling • Manufacturing and Industrial Engineering • Robotics and Control • Heat Transfer and Thermal • Power Plant Engineering • Risk and Reliability • Case studies and Critical reviews

The International Society of Ocean, Mechanical and Aerospace –science and engineering is inviting you to submit your manuscript(s) to [email protected] or [email protected] for publication. Our objective is to inform 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.3: January 2014

ISOMAse


Editors

Chief-in-Editor

Jaswar Koto (Universiti Teknologi Malaysia, Malaysia, Associate Editors

Adhy Prayitno (Universitas Riau, Indonesia) Agoes Priyanto (Universiti Teknologi Malaysia, Malaysia) Ahmad Fitriadhy (Universiti Malaysia Terengganu, Malaysia) Ahmad Zubaydi (Institut Teknologi Sepuluh Nopember, Indonesia) Buana Ma’ruf (Badan Pengkajian dan Penerapan Teknologi, Indonesia) Carlos Guedes Soares (Centre for Marine Technology and Engineering (CENTEC), University of Lisbon, Portugal) Dani Harmanto (University of Derby, UK) Iis Sopyan (International Islamic University Malaysia, Malaysia) Jamasri (Universitas Gadjah Mada, Indonesia) Mazlan Abdul Wahid (Universiti Teknologi Malaysia, Malaysia) Mohamed Kotb (Alexandria University, Egypt) Priyono Sutikno (Institut Teknologi Bandung, Indonesia) Sergey Antonenko (Far Eastern Federal University, Russia) Sunaryo (Universitas Indonesia, Indonesia) Tay Cho Jui (National University of Singapore, Singapore)

Published in Malaysia.

JOMAse

Jaswar Koto, P23-317, Dept. Aeronautical, Automotive and Ocean Engineering Faculty of Mechanical Engineering Universiti Teknologi Malaysia

Printed in Indonesia.

Teknik Mesin Faultas Teknik Universitas Riau, Indonesia

Journal of Ocean, Mechanical and Aerospace -Science and Engineering-, Vol.3

January 20, 2014

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

Crack Growth Simulation at Welded Part of LNG Tank

Kazuhiro Suga,a,* Takafumi Endo,b and Masanori Kikuchic

a)Mechanical Engineering, Tokyo University of Science,Suwa, Japan b)Mechanical Engineering, Faculty of Science and Technology, Graduate School of Tokyo University of Science, Japan c)Mechanical Engineering, Faculty of Science and Technology, Tokyo University of Science, Japan *Corresponding author: [email protected] Paper History Received: 26-December-2013 Received in revised form: 30-December-2013 Accepted: 29-December-2013 ABSTRACT The objective of this research is to evaluate crack growth behavior from a surface crack to a through crack in a liquefied natural gas (LNG) tank. Crack growth behavior is analyzed by superposition version FEM (S-FEM). As a crack grows, part of the local mesh crosses over the global mesh. In S-FEM, the local mesh is defined as existing in the global mesh. When a part of the local mesh crosses over the global mesh, the Young’s modulus of the part is made small to ignore the influence of the part. In this way, the stress distribution of the local mesh is improved. In this study, crack growth behavior under a tensile cyclic load is analyzed. After crack penetration, the crack shape becomes rectangular. Crack growth behavior under tensile bending loading is also analyzed by simulating the four-point bending test. In this simulation, the crack shape becomes trapezoidal after penetration. KEY WORDS: Fatigue Crack; S-version FEM; Surface Crack; Through Crack. 1.0 INTRODUCTION A tank storing liquefied natural gas (LNG) has many welded parts. So, in the design phase of the tank, it is necessary to prevent LNG leaking due to fatigue, which can generate a through crack. Crack growth behavior at the welded part under tensile cyclic loading or bending has been researched [1,2,3,4]. However, crack growth behavior at the welded part under tension and bending has not been researched. In this study, the superposition version finite

element method (S-FEM) [5,6,7] is used to analyze crack growth behavior under tensile bending loading.

In S-FEM simulation, the global mesh represents the whole structure and the local mesh represents a local part of the structure that has a crack. The displacement function of the local mesh is defined as existing in the global mesh. The case that the local mesh crosses over the global mesh has not been extensively investigated, although one study applied S-version FEM when a local mesh crosses over a global mesh [8].

This research investigates the effects of part of a local mesh crossing over a global mesh on the stress distribution and the stress intensity factor. To show the effects, Young’s modulus of the local part is set as E/100. Then, the crack growth process under tensile cyclic loading is simulated. Finally, the crack growth process from a surface crack to a through crack at a welded part under tensile bending loading is simulated. 2.0 NUMERICAL ANALYSIS To analyze crack growth behavior, S-version FEM is used. The stress intensity factor is calculated by the virtual crack closure-integral method (VCCM) [9]. 3.0 PROBLEM OF LOCAL MESH CROSSING OVER GLOBAL MESH 3.1 Conditions of numerical simulation When a part of the local mesh crosses over the global mesh, the effects on the stress distribution and stress intensity factor are studied. The simulation model is a plate, as shown in Figure 1. The material used in the analysis is aluminum alloy A2017. The thickness of the plate is 2.3 mm. In Figure 1, the left side is the model, and the right side is view from z axial direction. The area enclosed by the green line is the global mesh and that enclosed by the blue line is the local mesh. The red area indicates the initial


January 20, 2014


crack shape. The plate has a crack of crack depth 2.0 mm and aspect ratio 1.0.

Two cases are simulated by changing Young’s modulus in the local mesh that crosses over the global mesh. One case assumes E/100 as the Young's modulus. The other case uses E as the Young’s modulus. A cyclic load is applied to the plate, and the stress range is assumed to be 10 MPa.

10 [MPa]

18 [mm]2.3 [mm]

60 [mm]

y

xz

Global mesh

Local mesh

x

y

z

Figure 1: Geometry of plate 3.2 Analysis result The stress distributions are shown in Figure 2(a) and Figure 2(b). Figure 2(a) is the cross-sectional view of the cracked plane. In Figure 2(b), the x-axis is defined as the direction of the arrows in Figure 2(a). The y-axis is the stress of each direction at the Gauss point of blue area of Figure 2(a). When Young’s modulus is E, each stress distribution is negative near the boundary between the global mesh and the local mesh. However, by assuming a Young’s modulus of E/100, each stress distribution is approximately 0 MPa.

y

xz

Depth

2.3[mm]

Gauss point

Figure 2: (a) Cross section

-5

0

5

10

2 2.2 2.4 2.6

Stre

ss o

f x a

xial

dir

ectio

n [M

Pa]

Depth [mm]

-10

-5

0

5

2 2.2 2.4 2.6

Stre

ss o

f y a

xial

dir

ectio

n [M

Pa]

Depth [mm]

-10

0

10

20

30

2 2.2 2.4 2.6

Stre

ss o

f z a

xial

dir

ectio

n [M

Pa]

Depth [mm] EE

E/100S-FEM S-FEM

Figure 2: (b) Stress distribution The stress intensity factors of the two S-FEM cases and the global FEM are shown in Figure 3. The stress intensity factors of the FEM and S-FEM cases are close to each other.


January 20, 2014


0.0

0.2

0.4

0.6

0.8

0 30 60 90 120 150 180

Stre

ss in

tens

ity f

acto

rK I

[MPa

・m

1/2 ]

Eccentric angle [deg.]

EEE/100

θ

Eccentric angle θ

180 0

90

a : Crack depth2c : Crack length

a

2c

S-FEM S-FEM

Figure 3: Stress intensity factor

0.0

0.2

0.4

0.6

0.8

1.0

0 30 60 90 120 150 180

Erro

r [%

]

Ecccentric angle [deg.]EE

E/100S-FEM S-FEM

Figure 4: Errors of stress intensity factor between S-FEM and FEM Assuming that the stress intensity factor of FEM is correct, errors of the stress intensity factors between S-FEM and FEM are shown in Figure 4. The change of Young’s modulus does not affect the stress intensity factor, even if the local mesh crosses over the global mesh. So, even if the local mesh crosses over the global mesh, it is clear that the stress intensity factor is calculated accurately. By using this calculation method, crack growth behavior under tensile cyclic loading is analyzed. The crack growth behavior is also analyzed because the stress distribution is improved by assuming that the Young’s modulus is E/100. 4.0 CRACK GROWTH BEHAVIOR UNDER TENSILE CYCLIC LOADING 4.1 Numerical model The crack growth behavior from a surface crack to a through crack under tensile cyclic loading is analyzed. The numerical model is a plate like the one in Figure 5. The right figure is the cross-section view at the crack plane. The red line represents the initial crack. The initial crack shape is semi-elliptical, the crack length is 1.8 mm and the aspect ratio is 0.9. The plate is subjected to a tensile cyclic loading of 80 MPa. The material used in the

analysis is also aluminum alloy A2017. The crack growth magnitude is calculated by Paris’ law (equation (1)) [10]. In eq.(1), C is 1.314× 10-10 and n is 2.37 (unit:

mMPa ). In eq.(1), da is crack growth magnitude , dN is Number of cycles and ⊿KI is Stress intensity factor range. da / dN = C(⊿KI)n (1)

xy

z

240 [mm]

5 [mm]50 [mm]

80 MPa

y

x

z

Global mesh

Local mesh

Initial crack

Figure 5: Dimension of plate model 4.2 Result of numerical simulation Figure 6 shows the changes in crack shapes before penetration. As the crack grows, the crack growth rate increases. Figure 7 represents the changes in the stress intensity factor before penetration. When the number of cycles is 1.01 × 106, the difference of stress intensity factors near the surface becomes large.

Figure 6: Changes in crack shape before penetration


January 20, 2014


Cycle [cycle]00.361×106

0.706×106

1.01×106

θ

Eccentric angle θ

180 0

90

a : Crack depth2c : Crack length1

a

2c

Figure 7: Changes in stress intensity factor before penetration Figure 8 represents the relation between the aspect ratio and the crack depth. T is the thickness of the plate and a is the crack depth. As the crack grows, the aspect ratio decreases gradually and becomes approximately 0.8. This is the same result found in Lin’s [11] and Hosseini's [12] studies.

0.0

0.2

0.4

0.6

0.8

1.0

0.0 0.2 0.4 0.6 0.8 1.0

Asp

ect r

atio

[-]

a/t [-]Figure 8: Changes in aspect ratio

Based on the final crack shape before penetration, where the crack depth is 4.65 mm, the local model after penetration is generated. The crack shape is assumed to be generated after penetration. After crack penetration, the crack depth becomes 1.02 times the specimen thickness. This shape is used as the initial shape of the through crack. The number of cycles during this crack growth process is approximately 3.9×104 cycles, which is negligibly smaller than the total cycles from the initial crack to the crack depth 4.65 mm, which is 1.01×106 cycles (3.8%).

Figure 9 represents the changes in the through crack. The number of cycles is defined as 0 when the surface crack is transformed into the through crack. Each line is the crack shape measured every 3.6×104 cycles. As the crack grows, the through-crack shape becomes rectangular. Figure 10 represents the changes in the stress intensity factor for each crack shape. As the crack grows, the difference of the stress intensity factors becomes small. When the number of cycles is 1.28 × 105, all stress intensity factors are nearly the same. So, it seems that the crack grows while keeping the rectangular crack shape. This is the same

result of Nam’s study [13].

Figure 9: Changes in crack shape after penetration

Cycle [cycle]

0

0.362×105

0.713×105

1.28×105

0

5

10

15

20

25

30

0 1 2 3 4 5

Stre

ss in

tens

ity f

acto

rK I

[MPa

･m1/

2 ]

Crack depth [mm]Figure 10: Changes in stress intensity factors after penetration Figure 11 represents the relation between the crack length and the number of cycles. The lower side shows the through crack shape. where Crack length 1 and Crack length 2 are defined. As the crack grows, both crack lengths become similar. So, it is clear that the through crack grows and maintains the rectangular shape.

0

5

10

15

20

25

30

0 0.2 0.4 0.6 0.8 1 1.2 1.4

Cra

ck le

ngth

[mm

]

Number of cycles [×105cycle]

Crack length2

Crack length1

Crack length1

Crack length2

t

Figure 11: Relation between crack length and number of cycles

5.0 CRACK GROWTH BEHAVIOR AT WELDED PART 5.1 Numerical model Fatigue cracks are easily generated at welded parts [14]. So, the fatigue crack growth at the welded part is analyzed. To analyze crack growth behavior from a surface crack to a through crack at the welded part, the fatigue test is the four-point bending test shown in Figure 12. The thickness of the specimen is 12 mm and


January 20, 2014


the thickness of the added plate is 30 mm. By attaching the thick added plate to the specimen, the stress of the whole specimen becomes tensile. The experimental model is symmetric, so the global mesh is a one-half model. Figure 13 shows the global mesh. The red area is the shape of the initial crack, which is located at the welded part. The initial crack length is 5.6 mm and the initial crack depth is 1.2 mm. The welded part is a triangular shape in figure 13. The crack growth magnitude is calculated by Paris’ law (equation (1)). The material used in the analysis is aluminum alloy A5083. In eq.(1), C is 2.6×10-10 and n is 3 (unit: mMPa ).

Specimen200 [mm]

500 [mm]

30 [mm]

12 [mm]

Added plateWelded part

Figure 12: Four-point bending test

x

yz

180 [mm]204 [mm]

150 [mm]

12 [mm]

6 [mm]

x

z

y

Welded part

Figure 13: Dimensions of global mesh and location of the initial crack

5.2 Results of numerical simulation Changes in the crack shape are shown in Figure 14. Each line represents the crack shape every 6.6×104 cycles. Due to the crack growth, the crack growing rate increases. Figure 15 shows changes in the stress intensity factors due to crack growth.

The stress intensity factors are the highest near the specimen surface due to the stress concentration at the welded part and tensile bending loading. At 1.98×105 cycles, the stress intensity factors near the crack depth are high. At this time, the crack depth is 11.6 mm. So, stress concentrates at the crack tip and the stress intensity factors there become high.

Crack length [mm]

Cra

ck d

epth

[mm

]

Figure 14: Crack shape before penetration

0

2

4

6

8

10

12

0 30 60 90 120 150 180

Stre

ss in

tens

ity f

acto

rK

I[M

Pa･m

1/2 ]

Eccentric angle[deg.]

1.24×1050.664×105

1.98×105

0Cycle [cycle]

θ

Eccentric angle θ

180 0

90

a : crack depth2c : crack length1

a

2c

Figure 15: Stress intensity factors Based on the final crack shape before penetration, where the crack depth is 11.6 mm, the local model after penetration is generated. The crack shape is assumed to be generated after penetration, where the crack depth becomes 1.02 times the specimen thickness. This shape is used as the initial shape of the through crack. The number of cycles during this crack growth process is approximately 5.0×103 cycles, which is negligibly smaller (1.5%) than the total cycles from the initial crack depth to a crack depth of 11.6 mm, which is 1.98×105cycles.

By using the through crack, the crack growth behavior is analyzed after penetration. Changes in the crack shape are shown in Figure 16. The number of cycles is defined as 0 when the surface crack becomes the through crack. In Figure 16, each line is the crack shape every 1.03×104 cycles. As the crack grows, the crack shape becomes trapezoidal. Under cyclic tensile and bending loading, the through crack shape becomes rectangular. However, the crack shape becomes trapezoidal when the crack shape is affected by tensile bending loading.

Changes in the stress intensity factor are shown in Figure 17. Shortly after penetration, the stress intensity factors near the surface are high due to the stress concentration near the surface. As the crack grows, the difference in the stress intensity factor becomes small. In the following crack growth processes, the stress intensity factor increases in the whole crack front. It is noticed that the crack grows in a trapezoidal shape.

Crack length [mm]

Cra

ck

dept

h [m

m]

Figure 16: Crack shape after penetration


January 20, 2014


0

2.03×104

Cycle [cycle]

1.03×104

2.99×104

Crack depth [mm]

Stre

ss in

tens

ity fa

ctor

K I[M

Pa･m

1/2 ]

0

4

8

12

16

20

0 2 4 6 8 10 12

Figure 17: Stress intensity factor Figure 18 shows the relation between crack length and number of cycles. Crack length 1 and Crack length 2 are the crack lengths of the through crack. Shortly after penetration, the crack growth rate (gradient of the curve) of Crack length 2 is faster than that of Crack length 1. As the crack grows, the gradients of each curve become the same.

0

20

40

60

80

100

0 0.5 1 1.5 2 2.5 3 3.5

Crack length1

Crack length2

Cra

ck le

ngth

[mm

]

Number of cycles [×104 cycle]

Crack length1

Crack length2

t

Figure 18: Crack lengths of through crack Figure 19 shows the relation between the crack length1 and number of cycles from the surface crack to the through crack. As the crack grows, the crack growth rate becomes fast. After penetration, the crack growth rate becomes fast.

Cra

ck le

ngth

1 [m

m]

Number of cycles[×105 cycle]Figure 19: Crack growth rate

6.0 CONCLUSION It is found that when the local mesh crosses over the global mesh in the simulation by S-version FEM, the stress distributions and the stress intensity factors are hardly influenced. As a result, the through crack shape becomes rectangular under tensile cyclic load. The crack shape becomes trapezoidal under tensile bending loading after penetration. This is a reasonable result of the stress distribution. REFERENCES 1. Nakamura, T. (2006). Numerical Simulation of Fatigue

Crack Growth for Fillet Welded Joints, Conference Proceedings, the Japan Society of Naval Architects and Ocean Engineers, Vol. 3, pp. 355-358. (in Japanese)

2. Mori, T., Kyung, K.S., Miki, C. (1998). The Influence of Static Strength of Steel and Weld Material on Fatigue Strength of Cruciform Filet Welded Joints, Journal of Structural Mechanics and Earthquake Engineering, Vol. 605, No. I-45, pp. 289-293. (in Japanese)

3. Mitsui, Y., Kurobane, Y., Harada, K., Konomi, M. (1983). Fatigue Crack Growth Behaviors at the Toe of Fillet Welded Joints under Plane Bending Load, Journal of the Japan Welding Society, Vol. 52, No. 3, pp. 288-305. (in Japanese)

4. Nagai, K., Iwata, M., Sung-Won Kang, SOEWEIFY, (1981). A Consideration on the Fatigue Strength of Fillet Welded Cruciform Joints of Mild Steel under Completely Reversed Bending, Journal of the Society of Naval Architects of Japan, Vol. 149, pp. 260-267.


January 20, 2014


5. Fish, J., Markolefas, S., Guttal, R., Nayak, P. (1994). On Adaptive Multilevel Superposition of Finite Element Meshes, Applied Numerical Mathematics, Vol. 14, pp. 135-164.

6. Okada, H., Endoh, S., Kikuchi, M., (2005). Application of S-Version Finite Element Method to Two Dimensional Fracture Mechanics Problems, Transactions of the Japan Society of Mechanical Engineers Ser. A, Vol. 71, No. 704, pp. 677-684. (in Japanese)

7. Kikuchi, M., Wada, Y., Takahashi, M., Li Y. (2008). Fatigue Crack Growth Simulation Using S-version FEM, Transactions of the Japan Society of Mechanical Engineers Ser. A, Vol. 74, No. 742, pp. 812-818. (in Japanese)

8. Suzuki, K., Sando, A., Shimura, D., Ohtsubo, H., (2004). Study on the Local Area Exceeding Global Area in Mesh Superposition Method, Applied Mechanics, Vol. 7, pp. 383-389.

9. Okada, H., Higashi, M., Kikuchi, M., Fukui, Y., Kumazawa, M., (2005). Three dimensional Virtual Crack Closure-integral Method (VCCM) with Skewed and Non-symmetric Mesh Arrangement at the Crack Front, Engineering Fracture Mechanics, Vol. 72, pp. 1717-1737.

10. Paris, P.C., Erdogan, F. (1963). A Critical Analysis of Crack Propagation Laws, Journal of Basic Engineering, Trans. ASME, Ser. D, Vol. 85, pp. 528-535.

11. Lin, X.B., Smith, R.A. (1999). Finite Element Modeling of Fatigue Crack Growth of Surface Cracked Plates Part II: Crack Shape Change, Engineering Fracture Mechanics, Vol. 63, pp. 523-540.

12. Hosseini, A., MAHMOUD, M.A. (1985). Evaluation of Stress Intensity Factor and Fatigue Growth of Surface Cracks in Tension Plates, Engineering Fracture Mechanics, Vol. 22, pp. 957-974.

13. Nam, K.W., Matsui, K., Ando, K., Ogura, N. (1969). The Fatigue Life and Fatigue-Crack-Through-Thickness Behavior of a Surface-Cracked Plate (3rd Report, In the Case of Combined Tensile and Bending Stress), Transactions of the Japan Society of Mechanical Engineers Ser. A, Vol. 55, No. 51, pp. 1740-1747. (in Japanese)

14. Tateishi, K., Kyung, K.,S., Machida F., Miki, C. (1996). Influence of Welding Materials on Fatigue Strength of Fillet Welded Joints Made from High Strength Steel, Journal of Structural Mechanics and Earthquake Engineering, Vol.543, No. I-36, pp. 133-140. (in Japanese)


January 20, 2014


Study of the Effect of Low Profile Vortex Generators on Ship Viscous Resistance

Yasser M. Ahmeda, b,*, A. H. Elbatrana, H. M. Shabaraa

a)Faculty of Mechanical Engineering, UniversitiTeknologi Malaysia, 81310, UTM , Skudai, Johor, Malaysia b)Faculty of Engineering, Alexandria University, Egypt *Corresponding author: [email protected] Paper History Received: 18-December-2013 Received in revised form: 25-December-2013 Accepted: 29-December-2013 ABSTRACT A study of the effect of the well-known aerodynamic device low profile vortex generators (VGs) on the viscous resistance of the DTMB 5415 ship hull form through the control of the ship boundary layer separation is performed using the finite volume code Ansys CFX. The tetrahedral unstructural grids have been used for meshing the different cases. Different types of VGs have been tested, but the study has forced on two main types of VGs. The effects of VGs on the ship viscous resistance and its components have been investigated for the different cases in this study, and comparisons between the various results have been made. KEY WORDS: Vortex Generator; Viscous Resistance; Aerodynamic Device. NOMENCLATURE

εε 21 ,CC turbulence model constants

CF frictional resistance coefficient CPV viscous pressure resistance coefficient CV viscous resistance coefficient Gk turbulent kinetic energy due to mean velocity

gradient e device chord length

f device thickness or width g slots depth h device height k turbulent kinetic energy ui velocity components in the directions of xi

=(x,y,z)

i ju u′ ′ Reynolds stresses

VGXΔ distance between the VG trailing edge and the beginning of high pressure zone

z device spacing in vertical direction boundary layer thickness

ijδ Kronecker delta

ε dissipation rate of turbulence kinetic energy

tυ turbulent kinematic viscosity

εσσ ,k turbulence model constants

ρ fluid density

1.0 INTRODUCTION Flow separation control is a very important task for many industrial applications of fluid mechanics [1,2]. Controlling flow separation can result in a reduction of system resistance with consequence of energy conservation. Different methods can be used for controlling the flow separation, such as wall cooling, boundary layer suction or using vortex generators (VGs). There are numerous aerodynamic applications using VGs for reducing flow separation in internal and external flows.

Conventional passive VGs with 1≈δh have been used to control flow separation in the boundary layer by increasing the

δ


January 20, 2014


near wall momentum through the transfer of momentum from the free stream flow to the wall region. These VGs have been used to delay the separation of the boundary layer [3], to decrease resistance of aircrafts fuselages [4] and in other many applications. However, in some cases this type of VGs may cause a transferring of vehicle forward momentum into the unrecoverable turbulence in the wake region of the vehicle, which leads to an increase of the residual drag.

Kuethe [5] has investigated the effect of non-conventional VGs with δh of 0.27 and 0.42 on the boundary layer separation. These low profile VGs have the ability of reducing the area of velocity deficit in the wake region [5]. A lot of investigations and researches have been made to study the effect of the VGs height on the viscous flow control, and it has been found that the use of low profile VGs in the range of

5.01.0 ≤≤ δh can provide good flow separation control [6] with lower drag of the VGs.

In aeronautical applications the low profile VGs may take different shapes and sizes, but the most common one is the vane type, while in marine field Odelal [7] provides a study for a range of different VGs geometries used for hydraulic and marine equipments. Odelal shows that investigations should be made before choosing a specific configuration of VGs for a particular application.

Indeed, there are few research works in the marine field of using VGs to control boundary layer separation; therefore this study is an attempt to investigate the effect of low profile VGs on the ship viscous flow. The model of the DTMB 5415 unit [8] has been chosen for this study as an example of ship hull form, and different shapes and sizes of VGs have been tested in this study. The effects on the viscous pressure resistance and the frictional resistance have been investigated with each type of the VGs. 2.0 GOVERNING FLOW EQUATIONS The governing equations for the 3D steady incompressible turbulent flow adapted for this study are the continuity equation for mass conservation and Reynolds-Averaged Navier-Stokes equations for momentum transport. These equations can be written in Cartesian form as follows: Continuity equation

0i

i

ux∂

=∂

(1)

Momentum transport equation

( )1 ji ij i j

j i j j i j

uu upu u ux x x x x x

υρ

⎡ ⎤⎛ ⎞∂∂ ∂∂ ∂ ∂ ′ ′= − + + + −⎢ ⎥⎜ ⎟⎜ ⎟∂ ∂ ∂ ∂ ∂ ∂⎢ ⎥⎝ ⎠⎣ ⎦

(2)

Reynolds stresses 23

jii j t ij

j i

uuu u kx x

υ δ⎛ ⎞∂∂′ ′− = + −⎜ ⎟⎜ ⎟∂ ∂⎝ ⎠

(3)

Turbulence model equations

In this study the standard k ε− turbulence model and the renormalization-group (RNG) k ε− turbulence model [9] have been used for simulations. However, the best results have been obtained from using the standard k ε− turbulent model. The equations of the standard k ε− turbulence model can be expressed as follows:

ρεσμ

μρ −+⎥⎦

⎤⎢⎣

⎡

∂∂

⎟⎟⎠

⎞⎜⎜⎝

⎛+

∂∂

= kik

t

i

Gxk

xDtDk

(4)

and

( )k

CGk

CxxDt

Dk

i

t

i

2

21ερεε

σμ

μερ εεε

−+⎥⎦

⎤⎢⎣

⎡

∂∂

⎟⎟⎠

⎞⎜⎜⎝

⎛+

∂∂

=

(5)

3.0 DESCRIPTION OF DTMB 5415 HULL FORM AND VGs Model DTMB 5415 (Fig. 1) was conceived as a preliminary design for a Navy surface combatant ca. 1980. The hull geometry includes both a sonar dome and transom stern. Propulsion is provided through twin open-water propellers driven by shafts supported by struts.

The principal dimensions of the DTMB 5415 are listed in Table 1 and the tests conditions are listed in Table 2 correspond to the model tests from the David Taylor Model Basin (DTMB) in Washington D.C. and the Istituto Nazionale per Studied Esperienze di Architettura Navale (INSEAN) in Rome, Italy. Both computations and experiments are conducted in the bare hull condition.

Figure 1: The hull form of DTMB 5415 unit.

Table 1: Principal dimensions of DTMB 5415 [8].


January 20, 2014


Table 2: Test conditions for the Model scale [8].

In the current study vortex generators of delta shaped type

and groves of slots shapes inthe ship hull form have been chosen for the investigations (Fig. 2). Moreover, other types of VGs have been tested, but the results of the previously mentioned types have been the best. The basic dimensions of the VGs used for every run in this study can be seen in Table 3. In fact, the vortex generators have been chosen based on aerodynamic studies, hydrodynamics studies and structural requirements [6], [10, 11].

Figure 2(a) Geometry of delta shaped VGs, Figure 2(b) Geometry of hull slots

Table 3: Main specifications of the VGs used in this study.

Case case1a case1b

case2a case2b

case3a case3b

Case4a Case4b

case5a case5b

VGs type

–

Delta shaped vortex generators Slots

No. of VGs 4 4

h (m) 0.0375 0.01875 – g (m) – 0.0005 δh ~ 0.2 ~ 0.1 – he ~1.733 –

hf 0.133 – hz ~ 1.13 ~ 2.26 ~ 2.15 – ge – 20 gz – ~ 84

hX VGΔ ~ 1.2 ~ 2.4 – gXVGΔ – ~ 80

In fact, it was found from the numerical simulations

conducted in this research work that reducing or increasing the number of VGs than 4 did not give good results, so this number of VGs was chosen for introducing its results. The value of z/h for runs case4a and case4b is the average vertical distance ratio,

while in this case the VGs are placed at equal distances of 0.0939m on the hull surface. All VGs defined in the previous table were set normally to the hull form and located at the same distance from the beginning of the high pressure region on the hull form. However, the effect of varying the distance between the trailing edges of the VGs and the beginning of the high pressure zone ( hXVGΔ ) has been investigated in this study, as will be discussed latter. 4.0 COMPUTATIONAL DETAILS The computational domain has been chosen to be a quarter of an elliptical cylinder, as shown in Fig. 3. It stretches out for half the ship’s length in front of the ship hull, 1.5 times the ship’s length behind the hull, one ship’s length on the side and half the ship’s length under the still water surface. The main axis x, y, z, represent the streamwise, lateral and normaldirections, respectively, as shown in Fig. 3. The unstructural tetrahedral grids have been used for meshing the computational domain using ICEM CFD software. The grids are concentrated on the VGs and on hull form of the ship model, especially at bow and stern regions (Fig. 4). Different sizes of mesh grids have been used to study the impact of the grid sizes on the results. The same sizes of mesh elements and number of mesh layers have been used on the ship hull form for the different runs of coarsen grids and finer grids (more mesh layers have been used with runs of finer grids) to insure good comparisons between the different runs. The different number of grids used in this study can be seen in Table 4.

The discretisation process has been made by applying the finite volume method. The high resolution advection scheme [12] has been used for discretizing the convective terms to prevent solution divergent, and first order upwind advection scheme has been used for the turbulence equations. The pressure is interpolated using linear interpolation scheme, while the velocity is interpolated using the trilinear scheme [12].

Figure 3: CFX computational domain for DTMB 5415 unit. Figure 4 Numerical grids for the hull form of the DTMB 5415 unit. Left: the grids for the domain and the hull form without VGs (finer grids), right: the grids on the hull and VGs.


January 20, 2014


Table 4: Number of mesh elements used for every run.

Case No. of elements Case No. of

elements case1a 402280 case3b 1338943 case1b 653949 case4a 432644 case2a 603881 case4b 1340182 case2b 2327195 case5a 637030 case3a 432159 case5b 2363059

5.0 RESULTS AND COMPARISONS

Fig. 5 gives comparisons between the experimental and the numerical values of CPV, CF and CV of the different runs in this study, which can be considered as a first criterion for assessing the effect of the VGs on the ship viscous flow. Furthermore, Figs. 6 and 7 give comparisons between reduction or increase ratios in the ship viscous resistance components with respect to the resistance components of runs case1a and case1b due to the use of VGs, which give more clear indications about the effects of the VGs on the ship hull form viscous resistance.

Figure 5(a) CPV for coarsen grids runs

Figure 5(b) CPV for finer grids runs

Figure 5(c) CF for coarsen grids runs

Figure 5(d) CF for finer grids runs.

Figure 5(e) CV for coarsen grids runs.

Figure 5(f) CV for finer grids runs.

Fig. 6(a) results of hull

-5%

0%

5%

10%

15%

20%

25%

30%

case2a case3a case4a case5a

% R

educ

tion

%Cv

%CF

%Cpv

2,0E-04

4,0E-04

6,0E-04

8,0E-04

1,0E-03

exp case1a case2a case3a case4a case5a

CP

V

Hull Hull+VGs

3,0E-04

4,0E-04

5,0E-04

6,0E-04

7,0E-04

8,0E-04

9,0E-04

1,0E-03

exp case1b case2b case3b case4b case5b

CP

V

Hull Hull+VGs

2,86E-03

2,88E-03

2,90E-03

2,92E-03

2,94E-03

2,96E-03


CF

Hull Hull+VGs

2,86E-03

2,88E-03

2,90E-03

2,92E-03

2,94E-03

2,96E-03


CF

Hull Hull+VGs

3,16E-03

3,26E-03

3,36E-03

3,46E-03

3,56E-03

3,66E-03

3,76E-03

3,86E-03


CV

Hull Hull+VGs

3,16E-03

3,26E-03

3,36E-03

3,46E-03

3,56E-03

3,66E-03

3,76E-03

3,86E-03


CV

Hull Hull+VGs


January 20, 2014


Fig. 6(b) results of hull + VGs Figure 6: Percentage reduction in the value of CV and its components for runs of coarsen grids, referred to the values of runs case1a and case1b respectively.

Fig. 7(a) results of hull.

Fig. 7(b) results of hull + VGs. Figure 7 Percentage reduction in the value of CV and its components for runs of finer grids, referred to the values of runs case1a and case1b respectively.

It is well seen from Figs. 5(a) and 5(b) the reduction that occurred to the hull formCPV due to the effect of increasing the fluid velocity in the ship boundary layer originated from the effect of VGs. The figures also show the use of VGs with higher size ( 2.0≈δh ) may lead to higher reduction in the hull CPV, in comparison with the VGs of lower size. This off course refers to the strong vortices generated by these types of VGs. However, the effect of using VGs with higher size may lead to nearly very

small reduction in the value of total CPV (CPV of the hull + CPV of VGs), as shown in Fig. 6(b) (the reduction ratio is 2.47%), which makes the use of VGs with lower size ( 1.0≈δh ) is probably better. The runs of finer grids confirm this deduction, but show an adverse increase in the total value of CPV for run case2b, as shown in Fig. 7(b). This may refer to the complicated fluid structure generated in this case, which require more finer grids and may even more accurate turbulence model for investigating such case. In general, the use of VGs of slots type show very good effect in the value of the ship viscous pressure resistance and the total value of this resistance component with the coarsen and finer grids runs, as can be seen from Figs. 5, 6 and 7.

Fig. 8 shows the pressure distribution in the aft region of the ship hull for the cases with and without VGs. The figure shows the region of high pressure (low velocity region, where the flow separation occurs) is expanded in the aft direction due to the effect of VGs, which reduce the effect of the flow separation in this region and this in turn reduces the viscous pressure resistance of the hull form. However, the high pressure distribution around the VGs of the different runs shown in the previous figure indicates that the VGs themselves cause drag. The VGs of 2.0≈δh seems to affect greatly on the zone of high pressure in comparison with other types, but unfortunately due to their size they may lead to an increase in the overall viscous resistance which means more ship resistance at the end. Generally, the modification in the velocity distribution in the ship aft part as a result of using VGs may lead to an improvement in the axial flow distribution in the propeller wake region, which will lead in turn to an improvement in the propulsive efficiency.

Figure 8(a) Case1b (without VGs)Figure 8(b) Case2b (with VGs) Figure 8(c) Case3b (with VGs) Figure 8(d) Case4b (with VGs)

-5%

0%

5%

10%

15%

20%

25%

30%

case2a case3a case4a case5a

% R

educ

tion

%Cv%CF

%Cpv

-5%

0%

5%

10%

15%

20%

case2b case3b case4b case5b

% R

educ

tion

%Cv%CF

%Cpv

-30%

-25%

-20%

-15%

-10%

-5%

0%

5%

10%

15%

20%

case2b case3b case4b case5b

% R

educ

tion

%Cv

%CF

%Cpv


January 20, 2014


Figure 8(e) Case5b (with slots)

Figure 8: Pressure distribution on the ship aft part.

One of the main reasons for testing VGs with different sizes is the trial for choosing a device generates relatively strong vortices with small wetted surface area to reduce the effect of increasing the frictional resistance due to the addition of VGs. In general, the different devices used in this study did not lead to higher increase in the value of total frictional resistance coefficient, as can be seen from the results of the different runs in Figs. 5, 6 and 7, and the maximum increase in the value of CF was nearly 1.5% of VGs with 2.0≈δh which was expected due to their relatively higher wetted surface area.

The VGs of slots type show good reduction in the value of the total viscous resistance for the coarsen and finer grids runs and considered to be according to this study the best device with respect to the other types of VGs. Moreover, the VGs of

1.0≈δh show some improvement in the value of total CV as shown in Fig. 6(b), but this effect does not appear in the case of finer runs (Fig. 7(b)), which may refer to the insensitivity of the fluid solver for detecting the flow pattern correctly with finer grids runs. Figs. 5, 6 and 7 show that the results of runs case3a, case4a are approximately the same, which means the effect of the VGs on the boundary layer were nearly the same even with changing the distance between the devices. The same deduction can applied to the finer grids runs case3b and case4b.

Run case2b has been taken as an example for investigating the effect of varying the distance between the VGs and the beginning of high pressure zone, which represented by the parameter hX VGΔ . The values of 16.4 and 9.4 have been intended for hX VGΔ , in addition to the value of 2.4 of the original case. The effect of varying this parameter on the ship viscous resistance components and hence on the ship viscous resistance can be seen in Fig. 9 below. The results in the previous figure give the following indications:

I. The increase of hX VGΔ leads to lower increase in the

hull frictional resistance or its total value. II. The value of CPV of the hull and for the hull plus VGs

improved with the more decrease of hX VGΔ value. III. The viscous resistance is affected by varying hX VGΔ

which may become bigger than the original value of the hull form without VGs. This means the location of VGs need some kind of optimization to reach to the most

favorable value of hX VGΔ which lead to the better results.

Fig. 9(a) viscous resistance components for the hull form only, right Fig. 9(b): viscous resistance components for the hull from plus the resistance components of VGs. Figure 9: The effect of varying hX VGΔ parameter on the ship viscous resistance. 6.0 CONCLUSIONS

Numerical simulations for the viscous flow around a ship hull form with and without VGs using the commercial code Ansys CFX were carried out in this study. The obtained results showed generically the capability of the VGs on reducing ship viscous resistance component, but this may be accompanied by slightly increasing in the ship frictional resistance component. The resistance components of these VGs when added to the ship viscous resistance components this may lead to an increase in the total viscous resistance value. This give an important indication about the requirement of an optimization process for choosing the most suitable type of VGs required for a certain hull form. As well, it has been found that the final effect of the VGs on the ship viscous resistance will depend on: the type of VGs used, dimensions of VGs, number of VGs, the distance between the VGs trailing edges and the beginning of high pressure zone. However, experimental investigations are really

-10%

-5%

0%

5%

10%

% R

educ

tion

%Cv

%CF%Cpv

%Cv -1,34% 0,65% 0,57%

%CF 0,00% -0,21% -0,34%

%Cpv -8,51% 5,27% 5,44%

A B C

-20%

-15%

-10%

-5%

0%

5%

% R

educ

tion

%Cv

%CF

%Cpv

%Cv -2,34% -0,22% -0,30%

%CF -0,21% -0,34% -0,46%

%Cpv -13,74% 0,41% 0,59%

A B C


January 20, 2014


required in this field to obtain more clear picture and understanding about the actual effect of the VGs on the ship boundary layer separation and hence on the ship viscous resistance.

There were some differences between the results of the coarsen grids runs and the finer grids runs for detecting the effects of VGs on ship viscous resistance. In addition, the VGs of cylinder type has been tested also, but the results did not add to the results discussed in this paper due to the requirements of more investigations with this type of VGs, where the flow structure as well known is very complicated behind the cylinders at high Reynolds number. Generally, these may give indications of the requirement of further numerical investigations with: more finer grids, structural grids, transient simulations and more accurate turbulence models. ACKNOWLEDGEMENTS The authors would like to convey a great appreciation to the Faculties of Engineering in Universiti Teknologi Malaysia (UTM) and Alexandria University. REFERENCES 1. Gad-el-Hak M., Bushnell D. (1991), Separation control:

review, Journal of Fluids Engineering, Vol. 113, pp. 5-30. 2. Haines A.(1998), Know your flow: the key to better

prediction and successful innovation, AIAA Paper 98-0221, 36th AIAA Aerospace Science Meeting and Exhibit, Reno, NV, January 12-15.

3. Schubauer G., Spangenber W. (1960), Forced mixing in boundary layers, Journal of Fluids Mechanics, Vol. 8, pp. 10-32.

4. Calarese W., Grisler W., Gustsfson G. (1985), Afterbody drag reduction by vortex generators, AIAA Paper 85-0354, AIAA 23rd Aerospace Science Meeting, Reno, NV, January 14-17.

5. Kuethe A. (1972), Effect of streamwise vortices on wake properties associated with sound generation, Journal of Aircraft, Vol. 9, No. 10, pp.715-9.

6. John C. (2002), Review of research on low-profile vortex generators to control boundary layer separation, Progress in Aerospace Sciences, Vol. 38, pp. 389-420.

7. Oledal M. (1997), Application of vortex generators in ship propulsion system design, ODRA 97’ 2nd International Conference on Marine Technology, Computational Mechanics Publications, pp. 247-25.

8. WebPage:http://www.iihr.uiowa.edu/gothenburg2000/5415/combatant.html

9. Versteeg H., Malalasekera W. (1995), An Introduction to Computational Fluid Dynamics – The Finite Volume Method, Longman Group Ltd.

10. Godard G., Stanislas M. (2006), Control of decelerating boundary layer – part 1: optimization of passive vortex generators, Aerospace Science and Technology, Vol. 10, pp. 181-191.

11. Brandner P., Walker G. (2003) Hydrodynamic performance of a vortex generator, Experimental Thermal and Fluid

Science, Vol. 27, pp. 573-582. 12. CFX User’s Manual, Ansys CFX, 2013.


January 20, 2014


Design of a Body with Depth Control System for an Underwater Glider

Muhamad Fadli Ghani,a,* and Shahrum Shah Abdullah,b

a)Universiti Kuala Lumpur, Malaysian Institute of Marine Engineering Technology, Lumut, Perak b)UniversitiTeknologi Malaysia, Faculty of Electrical Engineering, Skudai, Johor *Corresponding author: [email protected] Paper History Received: 13-December-2013 Received in revised form: 20-December-2013 Accepted: 30-December-2013 ABSTRACT The underwater glider is used for deep water to observe large areas with minimal use of energy and move through the water by changing the body weight. The glider contained a cylindrical body attached with two wings and a fix tail. The controller has been designed to use a comparator circuit integrate with a pressure sensor to control the depth level. The pressure sensor mounted on the glider used to sense the underwater pressure. For the beginning, this underwater glider is limited to a depth of 0-10 meters. KEY WORDS: Underwater Gliders; Depth Control System; Cylindrical Body; Wings; Tail. 1.0 INTRODUCTION The concept of the underwater glider was proposed by Stommel (1989). Since 1995, US Navy Office of Naval Research has sponsored Autonomous Ocean Sensing Network (AOSN) program (Davis et al.2003) and have produced three oceangoing gliders including Slocum Glider (Webb et al. 2001) shown in Figure 1, the Spray Glider (Sherman et al. 2001) and the Sea Glider (Ericsen et al. 2001). These gliders are designed for long duration and ocean sensing missions (Osse et al. 2007).

Figure 1: Slocum Glider

Underwater gliders are a new class of Autonomous Underwater Vehicles (AUVs) (Rudnick et al. 2004) with fixed wings and a tail to glide through the ocean. They have many useful application such as in oceanographic sensing and data collection and also sea mapping. In this application, the gliders are very suitable because it’s capable for long duration missions, wide range areas, low cost and minimal used of power source. The gliders travel from place to place by produced of upwards and downwards glides (Webb et al. 2001). The glider glides in asaw tooth pattern by controlling their buoyancy using internal tanks and pumps. Propulsion of the gliders is created by changing the volume of the vehicle (Graver et al. 1998) either by moving oil from an internal tank to an external tank or by pumping seawater in or out of a tank. The glider has maintained a constant mass and changed its volume. Wings and body lift of the glider convert the vertical motion to a horizontal displacement like a saw tooth pattern


January 20, 2014


2.0 GLIDER DESCRIPTION The glider is designed and developed to operate as a platform for range of research in underwater technologies mainly relating deep water and long range period. Based on the Slocum Glider criteria, the design criteria for this project were that the glider should be:

• low cost material for the glider body • neutral buoyant • considered for maximum depth of 10 meters • mounted with pressure sensors as depth indicator

2.1 Hull Description The glider needs a pressure hull to store its components in a dry and watertight environment. The hull must allow the components to be easily accessible and maintainable in case of future changes or additions. The hull also needs to be corrosion resistant as it wills exposes to saltwater environment. Cylindrical hull provides the best structure and shape because spherical hulls offer the best structural integrity. 2.2 Design and Construction Process of Glider’s Body The process has been classified into two stages. First stage concentrates on the design concept of the glider. Therefore, computer-aided software such as the AutoCAD and MaxSurf are applied to sketch and animate the glider that are recommended and expected. The second stage expressed the fabrication development of glider body. The fiberglass is used as the main material for glider body fabrication. 3.0 ELECTRICAL /ELECTRONIC SYSTEM DESIGN 3.1 Sensor Suite The glider controller will receive feedbacks from the sensor which installed in the glider’s body for decision making. The glider controls its depth with feedback from a pressure sensor to activate the pump. The pressure sensor MPX4250 is used to measure the depth which produces small voltage when the depth increases. The MPX4250 needs supply voltage in between 4.85 to 5.35 Volts to operate. The MPX4250 is a low cost and capable to measure maximum pressure of 36.3psi or about 2.47atm. At sea level, pressure due to open air is 14.7psi or 1atm and for every 10meters of depth, the pressure increases about 1atm. The absolute pressure at 10meters underwater is 2atm or 29.4psi. 3.2 Power System The glider used battery pack that contains of 12 Volts for powering all electrical and electronic equipment. These batteries supply 5 Volts power lines to the sensor and 12 Volts power lines to the water pump. The onboard power supply is crucial to enable the glider to operate in autonomous mode. The battery pack is placed at the center gravity of glider’s body on the dry compartment to ensure the vehicle stability. 4.0 RESULTS The glider designed can be enhanced by using MaxSurf software shown in Figure 2. The software enables the user can design variety of dimension and can make any changes if required. Early

design, the glider was small scale of sizes, hence it changes the dimension because the material used needs a bigger size to be practically construct.

Figure 2: 3D MaxSurf

After several adjustments have been made to meets the

requirement, the construction of the glider begins with using fiberglass as the main material. At the first stage of construction, the plug is built by using polyvinyl chloride (PVC) pipe according to the hull size. The process to build the plug requires two weeks as it needs several adjustments to suits the design specification. Figure 3 shows the finishing product as the actual model of the glider.

Figure 3: Actual product

5.0 CONCLUSION This paper has been of tremendous success for my enhancement and advancement in knowledge and understanding of the robotic submarines. They are a part of the autonomous and unmanned vehicles used as low cost tools. The design structure and construction process of the project has been well discussed, while the research results and problem finding also has been describe. The practical body that we produce from our project will be used


January 20, 2014


to test the performance of depth control system and advanced control algorithm. With a limited budget, a practical body of underwater glider has been developed. ACKNOWLEDGEMENTS Author would very much like to express with grateful thank you to Universiti Teknologi Malaysia especially to higher management. The authors also would like to thank Universiti Kuala Lumpur for the conferences grant. REFERENCE 1. Davis, R. E., C. C. Eriksen and C. P. Jones. (2003),

Autonomous buoyancy-driven underwater glider. In: Technology and Application of Autonomous Underwater Vehicle, ed. G. Griffith, pp. 37- 58. Tayor and Francis.

2. Eriksen C.C.,T.J.Osse, R.D.Light, T.Wen, T.W.Lehman, P.L.Sabin, J.W.Ballard, and A.M.Chiodi. (2001), Seaglider: A long range autonomous underwater vehicle for oceanographic research, IEEE Journal of Oceanic Engineering, vol. 26, pp. 424-436.

3. Graver J., J.liu, C. Woolsey and N.E. Leonard. (1998), Design and analysis of an underwater vehicle for controlled gliding. Proc.32nd IEEE Conf. on Information Science and System, pp. 801-806.

4. Osse T.J. and C.C.Eriksen. (2007), The Deepglider: A Full Ocean Depth Glider for Oceanographic Research, IEEE Conference Oceans, pp.1-12.

5. Rudnick, D.L., R.E. Davis, C.C. Eriksen, D.M. Fratantoni and M.J. Perry. (2004), Underwater gliders for ocean research, Marine Technology Society Journal, 38(1): p. 48-59.

6. Sherman J.,R.E.Davis, W.B.Owens and J.Valdes. (2001), The autonomous underwater glider “Spray”, IEEE Journal of Oceanic Engineering, Vol. 26, pp. 437-336

7. Stommel.H. (1989), The Slocum Mission, Oceanography, Vol 2, no. 1, pp. 22-25.

8. Ueda, Y, and Rashed, SMH. (1990), Modern Method of Offshore Structures, Proc 1st Pacific/Asia Offshore MechSymp, Seoul, Vol 3, pp 315- 328.

9. Webb D. C., Simonetti P. J., and Jones C. P. (2001), SLOCUM, an underwater glider propelled by environmental energy, IEEE Journal of Oceanic Engineering, vol. 26, pp. 447- 452


January 20, 2014


Performance of VLCC Ship with Podded Propulsion System and Rudder

Jaswar Koto,a, b,* and Amirul Amin,a

a) Department of Aeronautical, Automotive and Ocean Engineering, Faculty of Mechanical Engineering, Universiti Teknologi Malaysia b) Ocean and Aerospace Engineering Research Institute, Indonesia c) Subsea Division,Technip Malaysia, Malaysia *Corresponding author: [email protected] Paper History Received: 20-December-2013 Received in revised form: 27-December-2013 Accepted: 1-January-2014 ABSTRACT Podded propulsion system becomes common installed in ships due to high maneuvering. The purpose of this paper is to discuss performance of both VLCC ships with podded propulsion and rudder. As initial offset data, published ship (SR221A) was used for generating hull form of a podded propulsion ship using Maxsurf software by maintaining the principal dimension: length, breadth, and draft. In order to suit installation of podded house, the stern part was modified. The hydrostatic data of both ships with podded propulsion and rudder are transferred to Ship Resistance and Propulsion Simulation software to determine the running speed by given same power. It was found that running speed produced by a ship using pod propulsion system lower than the speed produced by a ship using rudder. KEY WORDS: Rudder; Podded Propulsion System; Very Large Crude Oil Carrier. NOMENCLATURE RT Total craft drag

Friction resistance Wave making resistance Podded resistance due to podded house

Resistance due to podded body

Resistance due to podded strut Resistance due to podded fin

Resistance due to interference Resistance due to whirling

1.0 INTRODUCTION Propulsion is one of the important role should be considered during the ship design and construction because it will be affected on the ship manouvering, ship speed and the efficiency of the ship. Nowadays, researchers try to adapt the new technology of propulsor system that more efficient by using the electrical system such as podded propeller. The word POD means by propulsion with outboard electric motor. The podded propulsion being introduce since the middle of 1990. There are two company that mainly produce or manufactures the podded propulsion system which ABB and Rolls-Royce.

For the podded propulsion system, the propulsion and the steering system is combine together which means not using a rudder anymore to maneuver the ship. The podded propeller can be rotate at 360 degree of direction which improving the use of rudder. Podded propulsion system consists of the motor that connected directly to the podded propeller through the motor shaft. The podded propeller been using for variety types of ship such as ferries, cruise vessel, yachts, arctic tanker, drilling rigs, offshore supply vessels and icebreaker.

The most study nowadays is focus on the function or interaction of podded propeller as icebreaker which known as double acting tanker. Double acting tanker sail ahead at normal open sea and sail astern in ice. The main purpose of using double acting tanker is to reduce the cost that have been paid for using ice breaker when the tanker pass through the ice sea region such as at russia and arctic. The first double acting tanker be made is M.T. TEMPERA in 2002 and been propel by podded propeller.


January 20, 2014


This study focuses on the performance of the podded propulsion VLCC. The objectives of this study are to modify conventional hull design to be used for podded propulsion system, to determine resistance components of podded propulsion system, to determine speed of a podded propulsion system ship and compared with conventional ship at same BHP and RPM. 2.0 PODDED PROPULSION SYSTEM ON DOUBLE ACTING SHIP Early 1990’s, podded propulsion had been introduced as new electrical technology in marine and the originally concept was to improved the performance of icebreaker. Instead of using podded propulsion for icebreaker the naval architect had design a double acting ship hull that suitable for open water and ice condition by using the advantages of podded propulsion system.

The main role of the double acting tanker with the podded propulsion system is to go through the ice condition of sea. Without the double acting tanker, the vessel cannot break through the ice and without the podded propulsion system the vessel cannot propel in the stern position.

Therefore double acting tanker and the podded propulsion system are need each other to perform in ice condition and open water sea. Before this system are design, there are many of ways to overcome the problem during ice operation such as at Russian arctic but not reliable which are: • Using transhipment which uses different types of vessel for

different types of journey. • Using icebreaker to be assistance for tanker pass through the

ice. Figure 1: Double acting ship run head in open water and astern in ice condition [marine.com].

The traditional method of using icebreaker vessel as

assistance is not reliable because ship owner needs to spend extra money and its very costly rather than using DAT with podded propulsion system. DAT is very efficient in open sea and also in ice condition which perform ahead with open sea and astern in ice condition. There are some consideration should be taken during

design process of double acting tanker with podded propulsion which are stern shape, ice loads on propulsion and hull, behaviour in ballast condition and number of propeller. 3.0 NUMERICAL SIMULATION OF PODDED PROPULSION VLCC PERFORMANCE Simulation of podded propulsion ship performance was firstly hull and podded body and strut design using Maxsurf and then resistance and propulsion were estimated using Ship Resistance and Propulsion software as shown in Figure.2.

Figure 2: Flow chart of simulation of podded propulsion ship performance. 3.1 Hull Form Design The principal dimensions of podded propulsion VLCC as shown in Table 1 are maintained for both conventional and podded propulsion ships.

Table 1: Principal dimension of podded propulsion VLCC. Principle Dimension meter Length overall (LOA) 335.115 Length waterline (LWL) 325.528 Length between perpendicular (LPP) 320.00 Breadth moulded (Bm) 58.00 Depth moulded (Dm) 30.00 Draft 19.30

The design of hull was based on the existing design of SR221

as shown in Figure.3. There are 10 major stations for the offset but in the stern and the stem parts, the major stations were divided into small stations because to produce more accurate shape and produce smooth line which most of the stem and stern design are nonlinear lines. Accuracy and smooth of shape basically depend on number of waterlines and stations whereby more points to draw produce more accurate shapes.

Some modification on stern part is required to suit podded house installation as shown in Figure.4. The change of design

Hull Form Design

Podded House Design

Hydrostatic of ship

Resistance and Propulsion

Speed of ship


January 20, 2014


should be considered all the hydrodynamic characteristics of hull because it will effect on the powering calculation. For this research, there are 45 waterlines for each station which was added from the conventional design to make sure the enough draft and depth for podded house to be installed. The additional points can be determined by using mathematical method which is graphical method by derive the line equation and determine the points that required. The fairing hull form is shown in Figure.5. The hydrostatic data of the hull form was estimated using the Maxsurf software as shown in the Figure.6.

Figure 3: Stern hull design for conventional ship (rudder).

Figure 4: Stern hull design for podded propulsion system.

Figure 5: Hull form design of VLCC.

Figure 6: Input data of podded propulsion ship used in the Ship Resistance and Propulsion Simulation software. 3.2 Podded House Design Body and strut are main components of podded propulsion system. The design of these two components very important to make sure the system will produce more efficient power and less

resistance.The offset data of podded house was draw based on the actual podded propulsion system. The actual pod house has been scaled to the size required and sketch it using excel. Based on the design and shape, the modification should be done on the each point by using mathematical method to smooth it. The critical modification has done on the linear and polynomial region at front and back pod body shape. The design of pod body will affect on the resistance and flow of water that effect on powering estimation. Therefore, modification on the design should be done to make sure no extra power required because of form drag effect.

Strut profile uses the rudder’s offset of basis ship because of the shape of strut same as the shape of rudder but different in size. Basically the profile of the strut is called aerofoil shape which has standard shape. The different shape of aerofoil depends on the angle of attack and the chord length. The different between the aerofoil using in aircraft and the rudder of ship is rudder have same length of upper and lower chamber which is symmetry. The design of a podded house used in the present study was shown in the Figure.7.

Figure 7: Podded house design.

After modification, the pod design be drawn an attached to the

VLCC hull to estimate either the hull design suitable with the pod design or not. For improvement the design and shape of podded house, the podded house should be drawn and divided into small compartment and combine each compartment to be one shape of pod body. The podded house installed into the hull form is shown in the Figure.8.

Figure 8: Hull and podded propulsion system.

3.3 Resistance and Propulsion of Podded Propulsion Ship Basically, resistance of a ship with podded propulsion system ( ) is similar to other ships, but the only different is additional resistance due to podded house ( ). The total resistance of the podded propulsion ship ( ) can be written as


January 20, 2014


(1)

The podded resistance ( ) consists of resistances at body, strut, fin, due to interference and whirling as written below:

(2)

4.0 RESULTS AND DISCUSSION The resistance test is required to get the resistance data that very important to estimate the powering. There are many methods to determine the resistance data and for this study, using the numerical method to calculate and estimate the resistance parameter. The numerical method was run by the Naval-Offshore Simulation software which need input data such as principal dimension of the ship, offset data of design and hydrostatic data. Basically, the components that been calculated were the resistance of pod body, strut and hull.

In the simulation, a propeller with five blades of 8.9 meter of propeller diameter was selected. Based on the Figure.10, there are four major resistance component should be consider which are friction, delta, wave and additional resistance. Frictional resistance be the most high percentage of total resistance which about 86 percent and followed by correction resistance, wave resistance and additional resistance which are 8 percent, 3 percent and 3 percent respectively.

This study was focusing on the additional resistance which consists of the pod body, strut, interaction and whirling resistances. The fin resistance did not be considered because of there was no fin included in design the pod propulsion system for this study. From the Figure.9, the high percentage on additional resistance is strut resistance which about 77 percent. Then, body only provides about 20 percent of additional resistance due to high form drag rather than skin friction. Interaction resistance is resistance produce by the interaction between pod body and strut which about two percent. Next, about one percent is resistance produced by whirling effect which the resistance produce by vibration of the pod body and strut in the water.

Basically in design works, there are two conditions of propeller performances should be considered which are performance of propeller itself without effect of hull and the performance of propeller when installed behind hull. The performance of these two conditions actually gives different value of performance. The performance of the propeller behind hull is very important factors should be detail study because to show either the stern hull design will affect on the propeller performance or not which one of the main objective of this study.

The design of the propeller was based on the basis ship and from the basis data, the open water propeller performance can be calculated using the Naval-Offshore Simulation program. Also, the performance of the propeller behind of the hull can be calculated by using the offset data, principal dimension of hull and open water propeller performance data. This program uses numerical method to calculate the propeller performance behind the hull. Figure.10 shows the propeller efficiency of podded propulsion ship.

Figure 9: Resistance of a ship [Right] and component resistances of podded [left].


January 20, 2014


Figure.10: Estimated propeller efficiency of podded propulsion ship.

The efficiency of the podded propulsion system can be determined by doing the comparison of output data between the pod propulsion system and ship using rudder. The parameter that been used to compare was speed of the ship produced at same BHP and RPM conditions.

Based on the Figure.11 and Figure.12, at condition Maximum Continuous Rating (MCR) of engine there were different speed between conventional ship and ship using pod system. At same BHP and RPM, the speed of design ship produced by the conventional and the pod propulsion system was different which were 14.22 knot for the conventional and 13.97 knots for the pod propulsion ship. There are some losses due to resistance that make the ship slow in speed. But, there was some study on the resistance effect of form drag which because of pod shape. There are some losses due to resistance created by podded house and electrical losses. Other than that, based on the total area required for a ship using rudder and ship using pod propulsion system there were some different that effect on total resistance of both types of ship.

Table 2: Performance of ships with rudder and podded propulsion.

Type of ship BHP (kW) RPM Speed (knot)

Ship with rudder 20006 76 14.22 Ship with Podded Propulsion System 20006 76 13.97

Other than that, based on the total area required for a ship

using rudder and ship using pod propulsion system there were some different that effect on total resistance of both types of ship. Assume that the total area of the ship hull same for both ship but different in area of added components such as rudder, body and strut. Comparing with conventional system using rudder, the area of the podded house is higher about 30 %, as shown in Table.3, mostly from body.

Table.3: Total area of component Type of ship Components Total Area

(m2)Ship with rudder Rudder 240.578 Ship with Podded propulsion

Body and Strut 315.601

For ship resistance, basically, there are there major resistance

components should be considered which are friction, wave making and additional from podded system. It shows that 3 percent of total resistance contributed from podded propulsion. For podded resistance, five resistance components should be considered as follows: strut, body, fins, interaction between podded house and strut and whirling. The simulation shows that seventy seven percent of total resistance was contributed from the strut and twenty precent from the body.

Basically, for ship with rudder, there are only two components should be considered for resistance which are hull and rudder. But for the pod propulsion system, there are three main components should be considered which are strut, body and hull. Some additional resistances should be considered as well such as resistance produced by interaction between the strut and body. Therefore, generally podded propulsion system produces more resistance than using ship with rudder. Strut component contributes higher resistance than the body even though the total area of body is more than strut area which is 77 percent of total podded resistance as shown in the Figure.9. For this case of study, resistance divided into two main components which are frictional resistance and form drag resistances. Where, resistance of the body was mostly based on the form factor which shape of the body. But for total resistance, frictional resistance was mainly contributed resistance rather than form drag resistance. 5.0 CONCLUSIONS Based on the result and analysis of the study, the conclusions are obtained. Design of podded propulsion VLCC ship using Maxsurf has been done. There are some modification has been done on the stern hull design to suit it with the pod body and strut design. The resistance of the podded propulsion system has been divided into four components of resistance which are pod body, strut, interaction between them and whirling effect. Based on the result, the strut resistance contributes high percentage of additional resistance which about 77 percents and the pod body only contribute about 20 percent of additional resistance. The resistance components of podded house have been successfully calculated using Ship Resistance and Propulsion Simulation. Speed of the VLCC ship using podded propulsion system has been determined by using Naval-Offshore Simulation program and compared with the conventional ship using rudder. The speed produced by the ship using pod propulsion system lower than the speed produced by conventional ship using rudder which Vcon=14.22 knots and Vpod = 13.97 knots at same BHP and RPM condition.


January 20, 2014


Figure 11: Result of Pod propulsion system

Figure.12: Result of conventional ship using rudder.


January 20, 2014


REFERENCES 1. Alf Kare Adnanes,( 2003). Maritime Electrical Installation

and Diesel Electric Propulsion, ABB Technical Report, U.S.A.

2. Ali Dogrul, Yasemin Arikan, Fahri Celik (2010). A Numerical Investigation of Air Lubrication Effect on Ship Resistance, Macka Campus, Istanbul, Turkey.

3. C.D.Simonsen, (2000). Rudder, Propeller and Hull Interaction by RANS, PhD Thesis, Technical University of Denmark, Denmark.

4. F.R.Goldschmied, (1967). Integrated Hull Design, Boundary Layer Control, and Propulsion of Submerged Bodies, University of Utah, U.S.A.

5. H.C.Chen, S.K.lee, (2003). Chimera RANS Simulation of Propeller-Ship Interactions Including Crash-Astern Condition, ABS Techical Papers, U.S.A.

6. Jaswar and Yoshiho Ikeda (2002), Project Report to SHI: A Feasibility Study of Podded Propulsion Double Acting Oil Tanker, Osaka Prefecture University.

7. Jaswar (2005), Determination of Optimum Hull of Ice Ship Going, Proceeding of the 5th Osaka Colloqium, Japan.

8. Jaswar, A.Amin, Agustin, A.Maimun, Munirah, Mufti F.M, Jonathan and C.L.Siow (2012), Study on Powering of Podded Propulsion VLCC, The 6th Asia-Pacific Workshop on Marine Hydrodynamics, Johor, Malaysia.

9. J.Carlton, D.Radosavljevic, S.Whitworth, (2009). Rudder, Propeller, Hull Interaction : The Result of Some Recent Research, In-Service Problems and Their Solutions, 1st

International Symposium on Marine Propulsors, Trondheim, Norway.

10. Kimmo juurmaa, Tom Mattsson, Goran Wilkman Kvaerner Masa, (2001). The Development of The New Double Acting Ships for Ice Operation, Okhotsk Sea & Sea Ice, Mombetsu, Japan.

11. Lawrence J., (2009), A Study of The Resistance Characteristic of Surface Effect Ship Seals, University of New South Wales, Australia.

12. L. Bergh, U. Hellden, (2007). Electrical System in Pod Propulsion, Chalmers University of Technology, Goteborg, Sweden.

13. Luca Pivano. Experimental Validation of a Marine Propeller Thrust Estimation Scheme, Department of Marine Technology, Norwegian University of Science and Technology, Trondheim, Norway.

14. M.F.Islam, B.Veitch, P.Liu, (2007). Ecperimental Research on Marine Podded Propulsors, Memorial University of Newfoundland, St John’s, Canada.

15. M.F.Islam, B.Veitch, A.Akinturk, N.Bose, P.Liu, (2010). Performance Study of Podded Propulsor in Static Azimuthing Conditions, Oceanic Consulting Corporation, St . John’s, NL, Canada.

16. N.Sasaki, (2008). The Specialist Commitee on Azimuthing Podded Propulsion, 25th International Towing Tank Conferences, Fukuoka, Japan.

17. Robert d.moody,( 1996). Preliminary Power Prediction During Early Design Stages of Ship, University of Technology Cape Peninsula, South Africa.

18. S.Go, H.Seo, G.Choi, (2009). Study on the Powering Performance Evaluation for The Pod Propulsion Ships, Maritime Research Institute of Hyundai Heavy Industries

Co. Ltd, Ulsan, Korea. 19. S.Hanninen, M.Ojanen, A.Uuskallio, J.Vuorio, (2007).

Recent Developement of Podded Propulsion in Artic Shipping, Dalian University of Technoogy Press, China.

20. S.Bal, M.Guner, (2009). Performance Analysis of Podded Propulsor, Istanbul Technical University, Turkey.

21. Yoshitaka UKON (2008). Improvement of Propulsive Performance of a High-Speed Ship Equipped with a High-Powered Propeller, Kyushu University Research, Japan.

22. Z.Z.Rong, H.F.Wen, V.Kasilnikov, T.D.Hai. Progress in Analysis of Viscous Flow Around Podded Propulsor, China Ship Scientific Research Center, Jiangsu, China.

Head

d Office

ISOMAseResty MenaJalan Sising28282, PekaINDONESIhttp://www

e ara Hotel gamangaraja Nanbaru-Riau IA .isomase.org/

No.89

IISSN: 2354-70065

vol.3, january.2014 issn 2354-7065isomase.org/jomase/vol.3 jan 2014/vol-3.pdf · buana ma’ruf...

Documents