You are on page 1of 246

LONGITUDINAL STABILITY AND MOTION OF

TRIMARAN WING IN GROUND EFFECT MODEL


DURING TAKE-OFF

NOVERDO SAPUTRA

UNIVERSITI TEKNOLOGI MALAYSIA


PSZ 19:16 (Pind. 1/07)

UNIVERSITI TEKNOLOGI MALAYSIA


DECLARATION OF THESIS PROJECT PAPER AND COPYRIGHT

Authors full name : NOVERDO SAPUTRA

Date of birth : 21 OCTOBER 1983

Title : LONGITUDINAL STABILITY AND DYNAMIC MOTION


OF A TRIMARAN WING IN GROUND EFFECT (WIG) MODEL
DURING TAKE-OFF

Academic Session : SEM 2 2010/2011

I declare that this thesis is classified as:

CONFIDENTIAL (Contains confidential information under the Official Secret


Act 1972)*

RESTRICTED (Contains restricted information as specified by


the organization where research was done)*

OPEN ACCESS I agree that my thesis to be published as online open access


(full text)

I acknowledged that Universiti Teknologi Malaysia reserves the right as follows:

1. The thesis is the property of Universiti Teknologi Malaysia.


2. The Library of Universiti Teknologi Malaysia has the right to make copies for the purpose
of research only.
3. The Library has the right to make copies of the thesis for academic exchange.

Certified by:

SIGNATURE SIGNATURE OF SUPERVISOR

R 421182 PROFESSOR. DR. ADI MAIMUM BIN H. ABDUL MALIK


(NEW IC NO. /PASSPORT NO.) NAME OF SUPERVISOR

NOTES : * If the thesis is CONFIDENTAL or RESTRICTED, please attach with the letter from the
organization with period and reasons for confidentiality or restriction.
I hereby declare that I have read this thesis and in my opinion this

thesis is sufficient in terms of scope and quality for award of the

degree of Master of Engineering (Marine technology)

Signature :
Name : Professor. Dr. Adi Maimun Bin H. Abdul Malik
Date : September 2011
LONGITUDINAL STABILITY AND MOTION OF
TRIMARAN WING IN GROUND EFFECT MODEL
DURING TAKE-OFF

NOVERDO SAPUTRA

A thesis submitted in fulfillment of the


Requirements for the award of the degree of
Master of Engineering (Marine Technology)

Faculty of Mechanical Engineering


University Technology Malaysia

SEPTEMBER 2011
ii

DECLARATION

I declare that this thesis entitled Longitudinal Stability and Motion Wing in Ground
Effect model during take-off is the result of my own research except as cited in the
references. The thesis has not been accepted for any degree and is not concurrently
submitted in candidature of any other degree.

Signature :
Name : Noverdo Saputra
Date : September 2011
iii

DEDICATION

I dedicate this knowledge to:

My beloved father and mother


Development of science and technology
welfare of all people in the world

Nothing is impossible in this world; keep trying, learning and seeking blessing from
the God until the end of your life
iv

ACKNOWLEDGEMENTS

I would like to express special appreciation for my main supervisor Professor


Dr. Adi Maimun Bin H. Abdul Malik, for support, guidance, criticism and
suggestions during the study. I am also very grateful to my co-supervisor Dr. Agoes
Prijanto for advice, guidance and motivation. I will never forget and are very grateful
to my external supervisor, Professor Kirril V. Rohtdezvensky for basic concept and
knowledge about "ground effect". I am also very grateful to Dr. Maurizio Collu that
has been provided all references about hydrodynamic hull planning and his advice.
And also I am very grateful to Dr. John Rogers for vortex lattice algorithm and Mr.
Dingo Twedie for Savitsky Method Algorithm.

I would like to tank to the following individuals:

1 All Aerodynamic Laboratory staff for the assistance given during the wind
tunnel test.
2 All Marine Laboratory staff for the assistance given during the hydrodynamic
test.
3 All Centre of Composite Laboratory staff for the assistance given during
making model and prototype.
4 All Structure Laboratory staff for the assistance given during strength
assessment.
5 All my fellow colleagues in the Marine Laboratory, Saeed Jamee, Mobasser,
Ike Suharyanti. Mr. Rahimudin, Mr. Nurcholis, Mr.Yogi, Syaril.
v

6 All my fellow colleagues in Centre of Composite Laboratory, Mr. Imran, Mr.


Kamal, Mr. Rizal, Zamrie.
7 My entire PSM student, Rahim, Waqiyudin, Faisal, Tan Yong Seng, Syakur,
Izad.
8 My Best friend in KTC S47, Arief, Kazman,Uda Umar, Aaak, Yudis, Delin,
Mr.Prabowo, Akhry Nuddin
9 All contractor and supplier which has been supporting this research.
10 My family and my special friend for their encouragement during my study.
vi

ABSTRACT

Wing in Ground Effect is a relatively new concept of transportation


technology. It is more efficient than conventional aircraft and quicker compared to
conventional marine vehicles. However WIG is still not widely use as a public
transportation. One of the criteria to be is stability. Longitudinal stability of WIG
craft is still of concern to the designer and the solutions are being investigated.
Instability of a small WIG craft occurs when aerodynamic-hydrodynamic phase
changes into pure aerodynamic phase during the take-off. In this research,
investigations were conducted to determine the longitudinal static and dynamic
stability effect of Trimaran WIG craft during takeoff and to verify the factors
affecting its stability. Two parameters considered are aerodynamic and
hydrodynamic characteristics. The investigation resort vortex lattice method and
examines the effects of flat ground and end plate on the performance of aerodynamic
characteristic of the WIG craft. Planing hull has been chosen for the hull shape of the
WIG craft due to higher speed takeoff. The hydrodynamics of prismatic planing
surfaces, presented by Savitsky, is used to calculate the hydrodynamic characteristic.
Numerical result is compared to the experimental results and against published data.
The Static Stability Margin (SSM) for longitudinal static stability Trimaran WIG
model has been investigated and using the classical aircraft motion modification and
calculating the aerodynamic, hydrostatic and hydrodynamic forces, the complete
equation of motion that uses a small perturbation assumption for WIG during take-
off has been derived and solved. Finally, dynamic stability Trimaran WIG during
take-off has been investigated and analyzed using Routh-Hurwitz Stability Criterion
and Control Anticipation Parameter (CAP).
vii

ABSTRAK

Kenderaan Wing in Ground (WIG) merupakan konsep yang berbanding


baru dalam teknologi. Ia lebih cekap dari pesawat konvensional dan lebih pantas
berbanding kendaraan air konvensional. Walau bagaimana, WIG masih tidak boleh
digunapakai sebagai kenderaan pengangkutan awam. Salah satu kriteria yang perlu
dipenuhi adalah kestabilan. Kestabilan melintang kenderaan WIG masih menjadi
perhatian para pereka dan penyelesaianya sedang dikaji. Masalah kestabilan
kenderaan WIG yang kecil berlaku semasa pertukaran fasa aerodinamik-
hidrodinamik kepada fasa aerodinamik sepenuhnya ketika berlepas. Di dalam
penyelidikan ini, penyiasatan telah dijalankan untuk menentukan pengaruh kestabilan
melintang statik dan dinamik model knderaan trimaran WIG semasa berlepas dan
mengesahkan faktor yang mempengaruhinya. Dua parameter yang dipertimbangkan
ialah ciri aerodinamik dan ciri hidrodinamik. Kajian telah memilih kaedah kisi
vortex lattice dan memilih kesan tanah rata dengan hujung plat kepada prestasi ciri
aerodinamik dan ciri hidrodinamik kenderaan WIG. Planing hull telah dipilih
sebagai bentuk badan kenderaan trimaran WIG disebabkan perlunya kelajuan yang
tinggi untuk ia berlepas. Ciri Hidrodinamik permukaan prismatic untuk planing
hull yang dibentangkan oleh Savitsky telah digunakan untuk mengira ciri-ciri
hidrodinamik. Keputusan analisis berangka dibandingkan dengan keputusan uji
kajian atau dan data yang telah diterbitkan. Margin Kestabilan Statik (SSM) untuk
kestabilan static membujur model kenderaan trimaran WIG telah dikaji. Dan dengan
menggunakan mengambil kira modifikasi pergerakan asas pesawat udara dengan
mengira aerodinamik, hidrostatik dan daya hidrodinamik, persamaan lengkap
pergerakan yang menggunakan andaian perturbasi kecil, telah diterbitkan dan
diselesaikan untuk kenderaan WIG semasa ia berlepas. Akhirnya,kestabilan dinamik
Trimaran WIG semasa berlepas telah dikaji dan telah dianalisis menggunakan
Kriteria Stabiliti Routh-Hurwitz dan Faktor Kawalan Antisipasi (CAP).
viii

TABLE OF CONTENT

CHAPTER TITLE PAGE

DECLARATION ii

DEDICATION iii

ACKNOWLEDGEMENTS iv

ABSTRACT vi

ABSTRAK vii

TABLE OF CONTENT viii

LIST OF FIGURES xiii

LIST OF TABLES xviii

LIST OF ABBREVIATIONs xxiii

LIST OF SYMBOLS xxv

LIST OF APPENDICES xxxi


ix

1. INTRODUCTION 1
1.1 Background and problem statement 1

1.2 Purpose and objective of the study 3

1.3 Scope of this study 4

1.4 Significance of the study 4

1.5 Thesis outline 5

2. LITERATURE REVIEW 7
2.1 Historical ground effect vehicles 7

2.2 Aerodynamic WIG characteristic 10

2.3 Hydrodynamic WIG characteristic 11

2.4 Longitudinal static stability 16

2.5 Longitudinal dynamic stability 17

3, RESEARCH METHODOLOGY 21
3.1 Introduction 21

3.2 Initial design of model 21

3.3 Computational calculation 23

3.3.1 Vortex Lattice Method (VLM) 23

3.3.1.1 Preprocessor Vortex lattice Method


(VLM) 23

3.3.1.2 Solver Vortex lattice Method


(VLM) 24

3.3.1.3 Postprocessor Vortex lattice Method


(VLM) 25

3.3.2 Planing hull 25

3.4 Experimental work 26

3.5 Comparison and analysis of results 27


x

3.6 Longitudinal static stability & longitudinal dynamic


stability analysis 27

4, MATHEMATICAL MODEL DeVELOPMENT 29


4.1 Introduction 29

4.2 Vortex Lattice Method 29

4.3 Effect of flat ground and endplate on aerodynamic


characteristic 36

4.3.1 Effect flat ground 36

4.3.2 Effect endplate 37

4.4 Step by step Savitsky method 38

4.5 Longitudinal static stability 43

4.6 Longitudinal dynamic stability & dynamic motion 49

4.6.1 Aerodynamic force 49

4.6.2 Hydrodynamic force 50

4.6.3 Propulsion system, control system and


disturbance 51

4.6.4 Gravitational force 52

4.6.5 Equilibrium state, linearized, and cauchy


standard equation of motion 52

5. EXPERIMENTAL WORK 56
5.1 Wind Tunnel Test 56

5.1.1 Facility Low Speed Wind Tunnel Universiti


Technology Malaysia (LST -UTM) 57

5.1.2 Procedure wind tunnel test at LST-UTM 60

5.1.2.1 Preparation model 61

5.1.2.2 Set up Instrumentation 64

5.1.2.3 Testing 70
xi

5.1.2.4 Gathering data 72

5.1.2.5 Correction, normalization, and


linearization data 75

5.2 Free running test 77

6, RESULTS AND DISCUSSION 80


6.1 Validation Vortex Lattice Method (VLM) 80

6.1.1 Lift coefficient (CL) without endplate 81

6.1.2 5.1.2 Lift coefficient (CL) with endplate 88

6.1.3 Drag coefficient (CD) without endplate 94

6.1.4 Drag coefficient (CD) with endplate 101

6.2 Aerodynamic Trimaran WIG Model 108

6.2.1 Numerical Results 108

6.2.2 Experimental Results 113

6.2.3 Comparison numerical results with


experimental result 116

6.3 Hydrodynamic Trimaran WIG Model 123

6.3.1 Result and validation Savitsky Method (SM)


program with Hullspeed Maxsurf. 124

6.3.2 Free running test result 128

6.3.3 Comparison numerical result (SM) with free


running test 131

6.4 Longitudinal Static Stability Trimaran WIG Model 133

6.5 Longitudinal Dynamic Stability Trimaran WIG Model 137

6.6 Routh-Hurwitz Stability Criterion and Control


Anticipation Parameter (CAP) 140

6.7 Dynamic Motion Trimaran WIG Model 146

7. CONCLUSION AND RECOMMENDATION 148


xii

7.1 Conclusion 148

7.2 Recommendation 152

REFERENCES 154

APPENDIX A 159

APPENDIX B 163

APPENDIX C 170

APPENDIX D 177

APPENDIX E 189

APPENDIX F 191

APPENDIX G 196

APPENDIX H 210
xiii

LIST OF FIGURES

FIGURE NO. TITLE PAGE

2.1 Hull form 12


2.2 Photograph spray produced of a planing hull 13
2.3 Wetted area surface 14
2.4 Hydrostatic lift & hydrodynamics lift component 14
2.5 Dynamic airplane component 19
2.6 Dynamic WIG component 20
3.1 WIG trimaran Design 22
3.2 Flowchart preprocessor 24
3.3 Flowchart processor 24
3.4 Flowchart postprocessor 25
3.5 Flowchart savitsky method 26
3.6 Flowchart research 28
4.1 Flat wing sketch 30
4.2 Flat wing with dihedral sketch 30
4.3 Looking from upstream toward the trailing edge 31
4.4 Vortex filaments 32
4.5 New configuration SSM during take-off 48
5.1 Layout LST-UTM 59
5.2 LST-UTM main component 59
5.3 Aluminum block for model wind tunnel test 62
5.4 CNC milling MAHO 500 E2 62
xiv

5.5 Cold milling 63


5.6 Wing model for wind tunnel test 63
5.7 Set up fan power for wind speed 64
5.8 Honeycomb and screen LST-UTM 65
5.9 Heat exchanger LST-UTM 65
5.10 Balance and support system LST-UTM 66
5.11 3-Struts support 67
5.12 Street lantern light using single-strut support 67
5.13 Data Acquisition and Reduction System (DARS) LST-
UTM 68
5.14 Pressure measurement tools LST-UTM 69
5.15 Flow visualization LST-UTM 69
5.16 Pitot probe 70
5.17 Air stream check apparatus 71
5.18 Free running test schematic 78
5.19 Free running test 78
5.20 Recording data from running test 79
5.21 Real time data dashboard 79
6.1 Panel VLM 81
6.2 CL versus angle of attack for h/c= 0.1 and AR = 1 82
6.3 CL versus angle of attack for h/c= 0.1 and AR = 1.5 83
6.4 CL versus angle of attack for h/c= 0.1 and AR = 2 84
6.5 CL versus angle of attack for h/c= 0.3 and AR = 1 85
6.6 CL versus angle of attack for h/c= 0.3 and AR = 1.5 86
6.7 CL versus angle of attack for h/c= 0.3 and AR = 2 87
6.8 CL versus AOA for h/c= 0.1 and AR = 1, endplate he/c =
0.05 89
6.9 CL versus AOA for h/c= 0.1 and AR = 1.5, endplate he/c
= 0.1 90
6.10 CL versus AOA for h/c= 0.1 and AR = 2, endplate he/c =
0.05 91
6.11 CL versus AOA for h/c= 0.3 and AR = 1, endplate he/c =
0.05 92
xv

6.12 CL versus AOA for h/c= 0.3 and AR = 1.5, endplate he/c
= 0.05 93
6.13 CL versus AOA for h/c= 0.3 and AR = 2, endplate he/c
= 0.05 94
6.14 CD versus angle of attack for h/c= 0.1 and AR = 1 95
6.15 CD versus angle of attack for h/c= 0.1 and AR = 1.5 96
6.16 CD versus angle of attack for h/c= 0.1 and AR = 2 97
6.17 CD versus angle of attack for h/c= 0.3 and AR = 1 98
6.18 CD versus angle of attack for h/c= 0.3 and AR = 1.5 99
6.19 CD versus angle of attack for h/c= 0.3 and AR = 2 100
6.20 CD versus AOA for h/c= 0.1 and AR = 1, endplate he/c =
0.05 102
6.21 CD versus AOA for h/c= 0.1 and AR = 1.5, endplate he/c
= 0.05 103
6.22 CD versus AOA for h/c= 0.1 and AR = 2, endplate he/c =
0.05 104
6.23 CD versus AOA for h/c= 0.3 and AR = 1, endplate he/c =
0.05 105
6.24 CD versus AOA for h/c= 0.3 and AR = 1.5, endplate he/c
= 0.05 106
6.25 CD versus AOA for h/c= 0.3 and AR = 2, endplate he/c =
0.05 107
6.26 CL versus h/c every AOA for AR = 1.25 rectangular-
dihedral (50-50) wing trimaran WIG model without
endplate using VLM 109
6.27 CL versus h/c every AOA for AR = 1.25 rectangular-
dihedral (50-50) wing trimaran WIG model with
endplate 0.05c using VLM 110
6.28 CD versus h/c every AOA for AR = 1.25 rectangular-
dihedral (50-50) wing trimaran WIG model without
endplate using VLM 111
6.29 CD versus h/c every AOA for AR = 1.25 rectangular-
dihedral (50-50) wing trimaran WIG model with
endplate 0.05c using VLM 112
xvi

6.30 Experimental result of lift coefficient every angle of


attack (AOA) versus ground clearance (h/c) rectangular-
dihedral (50-50) wing trimaran WIG model with
endplate (AR = 1.25) 114
6.31 Experimental result of drag coefficient every angle of
attack (AOA) versus ground clearance (h/c) rectangular-
dihedral (50-50) wing trimaran WIG model with
endplate (AR = 1.25) 115
6.32 Comparison of CL versus AOA for h/c= 0.06 he/c=0.06
and AR = 1.25 between VLM, experimental LST-UTM,
and experimental LST-UTM after correction 117
6.33 Comparison of CL versus AOA for h/c= 0.1 he/c=0.06
and AR = 1.25 between VLM, experimental LST-UTM,
and experimental LST-UTM after correction 118
6.34 Comparison of CL versus AOA for h/c= 0.15 he/c=0.06
and AR = 1.25 between VLM, experimental LST-UTM,
and experimental LST-UTM after correction 119
6.35 Comparison of CD versus AOA for h/c= 0.15 he/c=0.06
and AR = 1.25 between VLM, experimental LST-UTM,
and experimental LST-UTM after correction 121
6.36 Comparison of CD versus AOA for h/c= 0.15 he/c=0.06
and AR = 1.25 between VLM, experimental LST-UTM,
and experimental LST-UTM after correction 122
6.37 Comparison of CD versus AOA for h/c= 0.15 he/c=0.06
and AR = 1.25 between VLM, experimental LST-UTM,
and experimental LST-UTM after correction 123
6.38 Hull without step and hull with step (Clementss step) 124
6.39 Planing hull simulation using Hullspeed [52] 125
6.40 SM & Hullspeed Maxsurf comparison [52] 125
6.41 Drag water resistance Trimaran WIG model with or
without step using SM 127
6.42 Lift coefficient planing surface with and without Step 127
6.43 Calibration RIG result (Thrust versus RPM) 129
6.44 Free running dashboard 130
xvii

6.45 Total drag trimaran WIG model result from free running
test 131
6.46 Comparison drag trimaran Wig model every speed
between numerical result (SM) and experimental work
(free running test) 132
6.47 Draft change during take off 132
6.48 SSM wing without endplate, h/c= 0.06 - 0.3 134
6.49 SSM wing with endplate, h/c= 0.06 - 0.3, he/c = 0.06c 135
6.50 SSM wing with endplate and tail, h/c= 0.06 - 0.3 he/c =
0.06c 136
6.51 Short Period Pitch Oscillation (SPPO) Trimaran WIG
every ground clearance (h/c) during take-off 137
6.52 Short Period Pitch Oscillation (SPPO) trimaran WIG
every angle of attack (AOA) during take-off 138
6.53 Long Period Pitch Oscillation (Phugoid) trimaran WIG
every ground clearance (h/c) during take-off 139
6.54 Long Period Pitch Oscillation (Phugoid) trimaran WIG
every angle of attack (AOA) during take-off 139
6.55 Comparison trimaran WIG during take-off between
Simulink Flightgear simulation and free running test 140
6.56 Pitch motion trimaran WIG model 146
6.57 Heave motion trimaran WIG model 147
xviii

LIST OF TABLES

TABLE NO. TITLE PAGE

3.1 Principle dimension of WIG trimaran 22


5.1 Model for wind tunnel test 61
6.1 Lift coefficient versus angle of attack for AR = 1, h/c =
0.1, based on VLM, CFD, and experimental by Jung et al 82
6.2 Lift coefficient versus angle of attack for AR = 1.5, h/c =
0.1, based on VLM, CFD, and experimental by Jung et al 83
6.3 Lift coefficient versus angle of attack for AR = 2, h/c =
0.1, based on VLM, CFD, and experimental by Jung et al 84
6.4 Lift coefficient versus angle of attack for AR = 1, h/c =
0.3, based on VLM, CFD, and experimental by Jung et al 85
6.5 Lift coefficient versus angle of attack for AR = 1.5, h/c =
0.3, based on VLM, CFD, and experimental by Jung et al 86
6.6 Lift coefficient versus angle of attack for AR = 2, h/c =
0.3, based on VLM, CFD, and experimental by Jung et al 87
6.7 Lift coefficient versus angle of attack for AR = 1, h/c =
0.1, endplate ratio he/c = 0.05, based on VLM, and
experimental by Jung et al 88
6.8 Lift coefficient versus angle of attack for AR = 1.5, h/c =
0.1, endplate ratio he/c = 0.05, based on VLM, and
experimental by Jung et al 89
xix

6.9 Lift coefficient versus angle of attack for AR = 2, h/c =


0.1, endplate ratio he/c = 0.05, based on VLM, and
experimental by Jung et al 90
6.10 Lift coefficient versus angle of attack for AR = 1, h/c =
0.3, endplate ratio he/c = 0.05, based on VLM, and
experimental by Jung et al 91
6.11 Lift coefficient versus angle of attack for AR = 1.5, h/c =
0.3, endplate ratio he/c = 0.05, based on VLM, and
experimental by Jung et al. 92
6.12 Lift coefficient versus angle of attack for AR = 2, h/c =
0.3, endplate ratio he/c = 0.05, based on VLM, and
experimental by Jung et al 93
6.13 Drag coefficient versus angle of attack for AR = 1, h/c =
0.1, based on VLM, CFD, and experimental by Jung et al 95
6.14 Drag coefficient versus angle of attack for AR = 1.5, h/c
= 0.1, based on VLM, CFD, and experimental by Jung et
al 96
6.15 Drag coefficient versus angle of attack for AR = 2, h/c =
0.1, based on VLM, CFD, and experimental by Jung et al 97
6.16 Drag coefficient versus angle of attack for AR = 1, h/c =
0.3, based on VLM, CFD, and experimental by Jung et
al. 98
6.17 Drag coefficient versus angle of attack for AR = 1.5, h/c
= 0.3, based on VLM, CFD, and experimental by Jung et
al 99
6.18 Drag coefficient versus angle of attack for AR = 2, h/c =
0.3, based on VLM, CFD, and experimental by Jung et
al. 100
6.19 Drag coefficient versus angle of attack for AR = 1, h/c =
0.1, endplate ratio he/c = 0.05, based on VLM, and
experimental by Jung et al 102
6.20 Drag coefficient versus angle of attack for AR = 1.5, h/c
= 0.1, endplate ratio he/c = 0.05, based on VLM, and
experimental by Jung et al 103
xx

6.21 Drag coefficient versus angle of attack for AR = 2, h/c =


0.1, endplate ratio he/c = 0.05, based on VLM, and
experimental by Jung et al 104
6.22 Drag coefficient versus angle of attack for AR = 1, h/c =
0.3, endplate ratio he/c = 0.05, based on VLM, and
experimental by Jung et al 105
6.23 Drag coefficient versus angle of attack for AR = 1.5, h/c
= 0.3, endplate ratio he/c = 0.05, based on VLM, and
experimental by Jung et al 106
6.24 Drag coefficient versus angle of attack for AR = 2, h/c =
0.3, endplate ratio he/c = 0.05, based on VLM, and
experimental by Jung et al 107
6.25 Lift coefficient every angle of attack (AOA) versus
ground clearance (h/c) rectangular-dihedral (50-50) wing
trimaran WIG model without endplate (AR = 1.25) using
VLM 109
6.26 Lift coefficient every angle of attack (AOA) versus
ground clearance (h/c) rectangular-dihedral (50-50) wing
trimaran WIG model (AR = 1.25) with endplate
(he/c=0.05) using VLM 110
6.27 Drag coefficient every angle of attack (AOA) versus
ground clearance (h/c) Rectangular-Dihedral (50-50)
wing trimaran WIG model (AR = 1.25) without endplate
using VLM 111
6.28 Drag coefficient every angle of attack (AOA) versus
ground clearance (h/c) rectangular-dihedral (50-50) wing
trimaran WIG model (AR = 1.25) with endplate
(he/c=0.05), using VLM 112
6.29 Experimental result of lift coefficient every angle of
attack (AOA) versus ground clearance (h/c) rectangular-
dihedral (50-50) Wing trimaran WIG model with
endplate (AR = 1.25) 114
6.30 Experimental result of drag coefficient every angle of
attack (AOA) versus ground clearance (h/c) rectangular-
xxi

dihedral (50-50) Wing trimaran WIG model with


endplate (AR = 1.25) 115
6.31 Lift coefficient versus angle of attack for AR = 1.25, h/c
= 0.06, endplate ratio he/c = 0.06, based on VLM,
experimental LST-UTM, and experimental LST-UTM
after correction. 117
6.32 Lift coefficient versus angle of attack for AR = 1.25, h/c
= 0.1, endplate ratio he/c = 0.06, based on VLM,
experimental LST-UTM, and experimental LST-UTM
after correction 118
6.33 Lift coefficient versus angle of attack for AR = 1.25, h/c
= 0.15, endplate ratio he/c = 0.06, based on VLM,
experimental LST-UTM, and experimental LST-UTM
after correction 119
6.34 Drag coefficient versus angle of attack for AR = 1.25,
h/c = 0.06, endplate ratio he/c = 0.06, based on VLM,
experimental LST-UTM, and experimental LST-UTM
after correction 120
6.35 Drag coefficient versus angle of attack for AR = 1.25,
h/c = 0.1, endplate ratio he/c = 0.06, based on VLM,
experimental LST-UTM, and experimental LST-UTM
after correction 121
6.36 Drag coefficient versus angle of attack for AR = 1.25,
h/c = 0.15, endplate ratio he/c = 0.06, based on VLM,
experimental LST-UTM, and experimental LST-UTM
after correction 122
6.37 Drag water resistance trimaran WIG model with and
without step 126
6.38 Calibration RIG result 128
6.39 Thrust from experiment free running test 130
6.40 Comparison drag trimaran WIG model between SM and
free running test 133
6.41 SSM on rectangular-dihedral reverse wing NACA 6409
A/R = 1.25 trimaran WIG model without endplate 134
xxii

6.42 SSM on rectangular-dihedral reverse wing NACA 6409


A/R = 1.25 trimaran WIG model with endplate (0.06c) 135
6.43 SSM on rectangular-dihedral reverse wing NACA 6409
A/R = 1.25 trimaran WIG model with endplate (0.06c)
and NACA = 0012 Tail A/R = 5 136
6.44 Routh-Hurwitz stability criterion for SPPO every ground
clearance (h/c) 141
6.45 Control Anticipation Parameter (CAP) for SPPO every
ground clearance (h/c) 141
6.46 Routh-Hurwitz stability criterion for Phugoid every
ground clearance (h/c) 142
6.47 Control Anticipation Parameter (CAP) for Phugoid every
ground clearance (h/c) 143
6.48 Routh-Hurwitz stability criterion for SPPO every angle
of attack (AOA) 143
6.49 Control Anticipation Parameter (CAP) for SPPO every
angle of attack (AOA) 144
6.50 Routh-Hurwitz stability criterion for Phugoid every
angle of attack (AOA) 145
6.51 Control Anticipation Parameter (CAP) for Phugoid every
angle of attack (AOA) 145
xxiii

LIST OF ABBREVIATIONS

WIG Wing In Ground Effect


CDGE Chord Dominated Ground Effect
SDGE Span Dominated Ground Effect
PIV Particle Image Velocimetry
LDA Laser Doppler Anemometry
RANS Reynolds Average Navier-Stokes
VLM Vortex Lattice Method
BEM Boundary Element Method
SPM Surface Panel Method
DHMTU Department of Hydro-Mechanics of the Marine Technical University
CFD Computational Fluid Dynamics
ACH Aerodynamic Center in Height
ACP Aerodynamic Center in Pitch
SSM Static Stability Margin
WISES Wing in Surface effect Ship
AOA Angle of Attack
LOA Length over all
BOA Breadth over all
B Hull breadth
LST-UTM Low Speed Wind Tunnel University technology Malaysia
ABB Allan Bradley
HCH Hydrodynamic Centre in Heave
AAMV Aerodynamically Alleviated Marine Vehicle
HCH Hydrodynamic Centre in Pitch
COG Centre of Gravity
SPPO Short Period Pitch Oscillation
xxiv

Phugoid Long Period Pitch Oscillation


CAP Control Anticipation Parameter
EDF Electric Ducting Propeller
NACA National Advisory Committee for Aeronautics
COG Centre Of Gravity
MATLAB language programming
FORTRAN language programming
vortex spinning, often turbulent, flow of fluid
chord long of the wing
span width of the wing
hull body of ship
fuselage body of airplane
tail small wing at
EHP small wing at
RULM Rectilinear Uniform Level Motion
xxv

LIST OF SYMBOLS

Aerodynamic

A aspect ratio ( b/c )


b wing span ( m )
c Chord length ( m )
CM moment coefficient (=L/0.5AU2 )
CL lift coefficient (=L/0.5AU2 )
CD drag coefficient (=D/0.5AU2 )
CDi induced drag coefficient
h/c ground clearance
he/c endplate ratio
N maximum number of element panel
cc cord along trailing leg of elemental panel (m)
angle of attack ( 0 )
dihedral angle ( 0 )
ground influence coefficient
endplate influence coefficient

sweep Angle ( 0 )
air density
vortex strength
F influence function geometry of single horshoe
S wing area (m2)
U free stream velocity (m/s)
u backwash velocity (m/s)
v sidewash velocity (m/s)
xxvi

w downwash velocity (m/s)


r1 , r2 vector distance
h/c ratio height ground into chord
h/b ratio height ground into span
a angle of attack
e downwash angle at the tail wing
z damping ratio
q pitch angle
n root of the speed subsidence mode
w frequency
alt lift curve slope of the tail wing
x,y,z distance along X,Y,Z
X ac distance center of aerodynamic from leading edge

X cg distance center of gravity from leading edge

Xh distance center of height from leading edge


body-axis system for plan form
wind axis system
distance along
x,y,z distance along X,Y,Z
X,Y,Z axis system for horsoe vortex
X ac distance center of aerodynamic from leading edge

X cg distance center of gravity from leading edge

Xh distance center of height from leading edge

XH distance center of hydrodynamic pitch from leading edge


X
derivatives height into surging
h
Z
derivatives height into heaving
h
M
derivatives height into pitching
h
Sref reference area
xxvii

Subscript

U backwash
V sidewash
w downwash
n index for elemental panel
H hydrodynamic pitch
OGE out ground effect
IGE in ground effect
E with endplate
WE without endplate

Hydrodynamic

LOA length over all


LWL length water line on (hydro)static condition
b chine beam (average) or beam of planing surface
measured between the chine( BPa)
displacement mass or vertical load on water
displacement volume
h vertical depth of trailing edge of boat (at keel)
below level water surface or depth of keel @ transom
LCG location of the longitudinal centre of gravity , forward of transom
VCG vertical centre of gravity, above baseline
p number of propellers
r number of rudders
Vs boat forward planing velocity or horizontal velocity of planing surface
b deadrise angle (degrees) - average, usually taken @ 0,5 Lp
e propeller shaft line inclination relative to the baseline (or keel line)
t trim (angle between planing bottom and horizontal)
CV speed coefficient
FN volume Froude number
xxviii

Cf Schoenherr frictional drag coefficient based on Reynolds number


Cf friction coefficient allowance for roughness of planing surface or
Correlation Allowance, Savitsky used CA = 0.0004 &
ITTC recommends CA = 0.003 for this friction line.( CA)
CL0 Lift coefficient @ zero deadrise
CL lift coefficient with deadrise surface
LP projected length of chine from transom to bow profile
LK Projected wetted keel length
LC Projected wetted chine length measured from transom to spray root
(stagnation line) intersection with chine (excluding spray)
LM Mean wetted length of pressure Area
BX beam max
or LM/b; mean wetted length / beam ratio
Effective increase in friction area length beam ratio
due to spray contribution to drag
LCP longitudinal location of centre of pressure
from trailing edge (i.e.transom)
CP Centre of pressure
N Resultant of pressure (hydrodynamic) and
buoyancy (hydrostatic) forces assumed acting normal to hull bottom
AP projected planing bottom area (excluding external spray strips)
or total bottom pressure area
Sw principal wetted surface area
(bounded by trailing edge, chines and heavy spray line)
SS area wetted by spray
f perpendicular distance off shaft line to Centre of Gravity (CG)
a the perpendicular distance between frictional drag-force
component Df and CG
fa Distance between Appendage Drag Da
(assumed as acting parallel to keel line) and CG
g acceleration due to gravity (or gravitational constant) = 9.81 m/s2
RN Reynolds number
n Kinematic viscosity of fluid (salt water @ 20 = 1*10-6)
xxix

VM Average water bottom velocity over the pressure area


q Angle between the keel (centreline) and
the outer edge of spray area measured in plane of bottom
specific weight of water (or mass density of water)
Df Frictional Drag-force component along bottom of surface
Da Appendage Drag (assumed as acting parallel to keel line)
T Propeller thrust along shaft line
d Diameter of shaft or bossing
c distance between N (pressure force applied to centre of pressure)
and CG measured longitudinally from transom stern and normal to N
RT total resistance

Dynamic Stability and motion

Fa Aerodynamic force

Fh Hydrodynamic force

Fc Control force

Fd disturbance force

Fg gravitational force

Fp Propulsion force

F0 equilibrium state of force

F' perturbation from datum


.
u surge displacement
.
q heave displacement, positive downward
.
w pitch rotation, positive bow up
A mass matrix
B damping matrix
C restoring matrix
D influence height into aerodynamic force
I moment inertia
xxx

Superscript
a related to aerodynamics
h related to hydrodynamics
perturbation value

Subscript
0 value at the equilibrium state
h derivative with respect to height

Experiment formula

Yw width of wake
q local dynamic pressure
q free stream
C mLE moment coefficient from leading edge

x(dc ) distance from leading edge (or other selected reference point)
P
pressure difference at each location
q
Pa pressure at inlet
Pb pressure at outlet
Y error every airspeed
xxxi

LIST OF APPENDICES

APPENDIX TITLE PAGE

A Trimaran WIG design 159


B Airfoil analysis 163
C Aero-hydro derivatives 170
D Vortex Lattice Method software 177
E Routh-Hurwitzh stability criterion code 189
F Savitsky method program 191
G Free running test set up 196
H Simulation program 210
CHAPTER 1

INTRODUCTION

1.1 Background and problem statement

Moving marine vehicles at high speed has been one of the biggest challenges
faced by naval architects and hydrodynamic researcher over the years especially after
the invention of aircraft, much thought have been given to find different methods that
can move ships quickly, Wing In Ground Effect (WIG) Craft is the most successful
one in terms of gaining high speed. The phenomenon of ground effect was observed
by many researchers since early in the birth of aviation. The advantages of using a
high speed craft in ground condition are commonly acknowledged by reduce drag
and increase lift. Wieselsberger (1921), Reid (1937) and Carter (1961),
hypothetically and experimentally was analyzed the influence of the ground on
aerodynamics wings. Analysis of experimental drag of wing with endplate shown
effect endplate in aerodynamic characteristic has investigated by Hemke (1927).
Absolutely, ground clearances and endplate ratio have influence in static stability
margin (SSM). Kumar (1968), Irodov (1970), Zhukov (1974), and Staufenbiel
(1988), Chun, and Chang (2002), all of them tell about the problem of longitudinal
stability in ground condition, where position aerodynamic centre in pitch (ACP) and
position aerodynamic in height (ACH) was influenced from the scenery of the
longitudinal stability every comparative position. Plentiful studies have been
conducted analyzing the influence of the ground effect on wing performance.
2

However, few largely disregarded. Rozhdestvensky (1998) presents a summation of


research investigating the influence of wing profile and platform on the positioning
of the two aerodynamic centers.

All problems above tell us about longitudinal static stability and longitudinal
dynamic stability during cruise which means that only aerodynamic affects the
stability. However, the stability problem that often occurs on the small WIG craft is
when WIGs phase change from hydrodynamic-aerodynamic phase into pure
aerodynamic phase during take-off. Collu et al (2010) has been tried to solve
advance mathematical framework for the longitudinal stability of a high-speed craft
with planing hull and aerodynamic surfaces. Eventually, a complete kinematics
model is been developed. Their observation illustrates a mathematical method for
performance Aerodynamically Alleviated Marine Vehicle (AAMV) in dynamic
condition. That vehicle was designed to take advantage of combination aerodynamic
forces and hydrodynamic force in high speed craft to get fuel efficiency and to reach
further and with a greater payload.

There are similarities between an AAMV with a WIG, when the WIG in
phase "aero-hydro" during take-off. Not only the aerodynamic force that worked
when she takeoff, but both of aerodynamic force and hydrodynamic force that
worked at that moment. In this research, observations were performed to determine
aerodynamic characteristic of NACA 6409 dihedral rectangular wing (50-50) of
aspect ratio 1-1.5 with taper 0.8. The observation were conducted using vortex lattice
method and investigating the influence of flat ground and endplate on enforcement
of a trimaran WIG for relative ground clearances of 0.01 < h/c < 0.2, with ratio
endplate 0.015< he/c <0.1 on angles of attack between 0 and 8. Planing hull has
been chosen for the Wig as high speed is necessary to takeoff .In 1964; a
comprehensive paper that summarized previous experimental studies on the
hydrodynamics of prismatic planing surfaces was presented by Savitsky. He
presented a method for application of these results for the design of moving ships.
Besides, many laboratories and research centers have conducted hydrodynamic
3

studies on several fundamental planing hull phenomena. All numerical result will be
validated with experimental results or other published work.

After that, the old concept Static Stability Margin (SSM) was modified by
adding hydrodynamic factor. Thus, SSM during takeoff will be presented with the
new configuration with three criteria, first, the position Aerodynamic centre in
Pitch (ACP) should be located downstream of the position Aerodynamic Centre in
Height (ACH) , second, the position of center of gravity (COG) of the craft should
be located upstream of the aerodynamic center of pitch (ACP) and third, the
position Aerodynamic Centre in Height (ACH) should be located upstream of
Hydrodynamic center in Pitch (HCP). The classical aircraft motion has been
modified by calculating the aerodynamic force, hydrostatic and hydrodynamic
forces, using a small perturbation assumption the full equations of motion WIG
during takeoff are derived and solved.

1.2 Purpose and objective of the study

The major aim of this investigate is to overpass this problem by investigating


a new configuration equation of motion WIG by calculating the equal importance of
aerodynamic and hydrodynamic forces in small-disturbance composition. The
arithmetical model of this framework was developed to investigate the longitudinal
static stability and longitudinal dynamic stability of a trimaran WIG during take-off.
4

1.3 Scope of this study

i. To estimate aerodynamic characteristic trimaran WIG using vortex lattice


method with criteria; NACA 6409 dihedral rectangular wing, aspect ratio
1.25, taper 0.8.
ii. To investigate influence of flat ground and endplate on the
performance of a trimaran WIG for several ground clearances, with ratio
endplate 0.06 on angles of attack between 0 and 8.
iii. To calculate hydrodynamic characteristic of prismatic planing surfaces
using Savitsky method.
iv. To validated the numerical results by experiment or publish work or
commercial software.
v. To arrange Static Stability Margin (SSM) criteria on WIG trimaran for
investigate longitudinal static stability WIG trimaran during takeoff.
vi. To arrange configuration equation of motion by calculating the equal
importance of aerodynamic and hydrodynamic forces in small-
disturbance composition for investigate longitudinal dynamic stability
WIG trimaran during takeoff.

1.4 Significance of the study

The significance of this study is to investigate longitudinal static stability and


dynamic stability of a trimaran WIG model during take-off. Where we should be to
calculate aerodynamics characteristic and hydrodynamics characteristic using
numerical equation and compare these calculation with experimental data or
commercial software.
5

1.5 Thesis outline

This thesis is managed into seven chapters. There are:

The first chapter provides about the background and problem statement of the
study, purpose and objective of the study, scope of the study, significant of the study.
Finally, thesis managing is presented.

The second chapter tells about literature review; history of vehicles on ground
condition, aerodynamic characteristic of vehicles on ground condition,
hydrodynamic of ground effect vehicles, longitudinal static stability of ground effect
vehicles, and dynamic stability of ground effect vehicles.

Chapter three provides the research methodology of longitudinal stability and


dynamic motion WIG trimaran during takeoff, where the stages of research are initial
design model, computational calculation, experimental work, comparison and
analysis of computational results with experiments or other published researchers,
longitudinal static and dynamic stability.

Chapter four propose mathematical modeling of aerodynamic characteristic


wing in ground effect using vortex lattice methods, effect ground effect on lift
coefficient, effect endplate on lift coefficient, hydrodynamic wing in ground effect
using Savitsky methods, analysis static stability ground effect with new configuration
static stability margin, and last on this chapter try propose new configuration
equation motion of a trimaran wing in ground effect during take-off for analysis
Routh-Hurwitz criteria and Criteria for public transportation(CPT).
6

Chapter five provides procedure experiment in wind tunnel test at Low Speed
Tunnel (LST) UTM and free running test.

Chapter six provides comparison results of computational calculation and


experimental, also analytical results

Chapter seven presents the conclusion and suggestion which can be used for
further research.
CHAPTER 2

LITERATURE REVIEW

2.1 Historical ground effect vehicles

The miracle of ground effect was attracted many researchers to observed


since early in the birth of aviation. The advantages of using a high speed craft in
ground condition are commonly acknowledged by reduce drag and increase lift
(Rozhdestvensky 2000). These observations began in the early 1920 Prandtl (1920)
has tried to make theory of lifting surface. In 1921, the revolutionary ground effect
scientist, Wieselsberger (1920) on wing resistance near the ground presents a
theoretical and experimental understanding to explain the performance of the planar
wing in ground effect, and find the cause of the increasing L/D ratio. And find ways
to reduce ground effect by finding a suitable geometry wing.

Around 1920, for the first time this effect was described and was carried out
in this field. The seaplane Dornier DO-X could cross the Atlantic Ocean by flying
low over the waves of the sea. This concept is intended to save fuel when they should
return to his base with a damaged engine condition during the Second World War
(SE 1996, Korolyov 2007).
8

After seeing some of the potential of the ground effect, some countries began
to seriously entered research and develop this effect to be used in the aviation world,
with the low flying can reduce fuel usage. In 1935, Toivo J. Kaario, try to develop
the Aerosledge No.8, but did not receive enough funding for further development (SE
1996, Rozhdestvensky 2000, Rozhdestvensky 2006,Chun 2008).

In 1963, the KM Caspian Sea Monster most famous ekranoplan was built by
Russia with displacement 500 ton (Halloran 1999). Also in 1963 was developed the
small single seat Collins X-112 aerofoil boat with low aspect ratio reverse delta
wing, dihedral and forward sweep, designed by German Scientist Alexander
Lippisch and built in America (SE 1996, Halloran 1999, and Leon 2007).

In 1967, China Ship Scientific Research Centre (CSSRC) developed various


WIG XTW series, which are based on a wing- tail configuration and Lippisch main
wing designs, the variant ever built are XTW-I, XTW-II, XTW-III, and XTW-IV (SE
1996, Leon 2007).

In 1979, the Soviet was built A-90 Orlyonok. Designed in 1974, only five
was built from the proposed 120, and were designed to quickly transport with larger
payloads (SE 1996, Leon 2007).

In1980s MARIC was developed AWIG (Amphibious WIG) vehicles.


AWIG-750 was built after testing approximately 80 radio controlled model. The
AWIG-750 has thrust 745kg with 4 internal combustion engines and 2 propellers (SE
1996, Leon 2007).
9

In 1987, the Lun-class Ekranoplan was proposed by the USSR. This craft was
equipped with missile launchers for anti submarine warfare. The MD-160, one kind
of this type was trialed for three years but was never used in operation (Leon 2007).

In 1990, the Aerocon Dash 1.6, large wingship was built by the US Company
Aerocon. This vehicle has thrust 5000 tons and is capable of a cruise speed of 400
knots (740km/h) (Rozhdestvensky 2006, Leon 2007). In thats year, Pelican was
developed by Boeing Phantom Works. The craft has capability to fly in ground effect
with altitude 20ft but also can fly over ground effects with altitude around 20,000 ft,
with span twice that of the worlds largest aircraft (the An-225), 0.4 Ha (SE 1996,
Leon 2007).

At the last 1990, Airfish-3 for four seat model was tested at speeds of
120km/h and was able to reach a range of 370 km. This vehicle was designed to
operate in ground effect with maximum dynamic jumps around 4.5m. After
successfully launching thats product, launched a new product Flightship 8 (FS-8)
for eight seat people and 2 crew members (Rozhdestvensky 2006).

In 1995, China Ship Scientific Research Centre (CSSRC) developed AWIG-


751, under the name Swan-I (2006, 2007). In 1995, Russia again raises new product
WIG, the Amphistar Aquaglide series has developed by the company Technology
and Transport with maximum take-off weight is 2720 kg, and cruising speed around
150km/h - 450km/h (Rozhdestvensky 2006, Leon 2007).

WIG developments not only in developed countries only, but travels to


developing countries, Indonesia have been success launch two products. Belibis NA-
5 & Belibis NA-6 successfully tested fly at Jatiluhur Reservoir, West Java. This craft
10

has designed for eight seater and cruising speed of approximately 144 km/h (BPPT
2007).

2.2 Aerodynamic WIG characteristic

Since the birth of aviation, many researchers have successfully developed


WIG crafts that fly by the water surface. The development of WIG vehicles is
currently going on in many countries. The potential fuel savings and speed
advantages over other modes of water transportation are the stimulus. Ludwig
Prandt'l (1920) on assumption that the lift is distributed in half an ellipse over the
wing spans shown in Theory of Lifting Surface. The ratio of the distance of the point
considered to the wing span is represented by h/b, and then the term is more familiar
with Span Dominated Ground Effect (SDGE).

The theoretical treatment of ground effect presented by Wieselsberger (1921)


for predicting the reduction in induced drag for wing at various heights of the quarter
chord of the wing above the ground (h/c), then the term is more familiar with Chord
Dominated Ground Effect (CDGE) Experiments have been full field by Reid (1937)
and Carter (1961).

The frictional drag of the endplates can reduce induced drag, is sufficiently
large to increase the efficiency of the wing. Analysis of experimental drag of wing
with endplate has done by Hemke (1927).
11

Margason and Lamar (1971) has done made Vortex Lattice Method (VLM)
program for calculate subsonic aerodynamic characteristic of complex planforms
using FORTRAN.

Zhang et al. (2002) reported the influence of tip vortex characteristics on the
aerodynamic performance of a cambered airfoil. Tip vortices generated by a
cambered, single element wing operating in ground effect were studied using a range
of methods including Particle Image Velocimetry (PIV), Laser Doppler Anemometry
(LDA), surface pressures, force balance measurements, and surface flow
visualization. They showed vortex breakdown occurred as the wing was lowered to
the ground, leading to a slowdown in the force enhancement. They equipped wing
with end plates and operated in ground effect.

Chun and Chang (2003) analyzed turbulent flow around two-dimensional


wings in ground effect with incompressible Reynolds Average Navier-Stokes
(RANS) equations which are approximated by finite difference schemes. The main
objective of their study was to clarify the two-dimensional ground effect and its flow
characteristics due to different ground boundary conditions, moving and fixed
bottom boundary. According to their computational results, the difference in the lift
and moment simulated by two bottom conditions is negligible, but the drag force
simulated by the fixed bottom is to some extent smaller than that by the moving one.

2.3 Hydrodynamic WIG characteristic

One of the major difficulties in designing a WIG is to find a method to reduce


drag during takeoff. As a WIG needs to run at high speed to lift itself from the
surface of the water, it faces significant amount of hull resistance, so that it can affect
12

the dynamic motion on a wig .So to get a solution of this problem (reducing drag) is
a worthy effort (Noverdo 2010). There are some option hulls forms can be used for
WIG craft. Research and development of advanced hull forms is a continuous to
developed, though not necessarily for reduced drag at high speed craft. Because of
the high speed requirements of WIG craft, displacement hulls are poor candidates, as
hull resistance is much too high (Todd 2002). Figure 2.1 shows the large range of
high-speed hull forms.

Figure 2.1: Hull form (Todd 2000)

Planing hull has been chosen for the Wig as high speed is necessary to take-
off .In 1964 a comprehensive paper that summarized previous experimental studies
on the hydrodynamics of prismatic planing surfaces was presented by Savitsky
(1964). He presented a method for application of these results for the design of
moving ships. Besides, many laboratories and research centers have conducted
hydrodynamic studies on several fundamental planing hull phenomena. The
underlying principles of high speed planing craft resistance have been treated by
DuCane, Clyton and Bishop. Viscosity and free surface effects, including spray and
overturning waves, play significant roles, making both experimental and numerical
13

predictions of suitable hull parameters very difficult. The planing hull is a kind of
hull that is specifically designed for a craft to achieve relatively high speed on the
surface of water. A planing hull makes the water to be pushed down and to the sides
as the hull moves forward. By the wake depression behind the hull, the downward
motion of the water is observed. The pushing of the water sideways can be observed
by the spray produced to the sides of the hull. Figure 2.2 show downward and
sideways movement of water builds pressure under the hull (Saunders 1957).

Figure 2.2: Photograph spray produced of a planing hull (Saunders 0)

The total wetted area surface of a planing hull is actually divided into 2
regions (Figure 2.3), the pressure area and spray area. The pressure area is the load
carrying area of the planing hull. The spray area contributes to the drag, but is not
considered to support any portion of the load (Savitsky 1964). The configuration will
be present as follows by (Todd 2002).
14

Figure 2.3: Wetted area surface (Savitsky 1964)

Planing hull significantly reducing wetted surface area, whereas displacement


hulls have longitudinal and transverse curvature, the planing hull has a transverse
deadrise section and straight buttock lines to induce early flow separation. The
negative dynamic pressure induced on the convex hull surfaces causes a large trim by
the stern, increasing resistance when a traditional displacement hull operates at high
speeds. The planing hull is designed to develop positive dynamic pressure, so the
displacement decreases with increasing speed (Todd 2002).

Figure 2.4: Hydrostatic lift & hydrodynamics lift component (Savitsky 1964)
15

The investigation hydrodynamic planing hull has been widely applied to the
researchers. Many methods have been used to estimate. Hassan Ghassemi, Mahmoud
Ghiasi, has been tried to combine Boundary Element Method (BEM) for predicted
induced drag, the frictional drag (the boundary layer theory), and the spray drag
(practical method) .All combination has been used to determine hydrodynamic force
of planing hull in the calm water condition. In this investigation was investigated by
an amount of elements with steady speed latent every element, where induced force
was investigated using the free surface elevation condition and the Kutta condition at
the transom stern. And induced drag and lift force was investigated by intended
distributions of dynamic pressure (see Figure 2.4). The frictional resistance was
investigated using the boundary layer analysis method, based on estimation of the
momentum integral equation. The area of spray has present by a particular practical
approach (Ghassemi 2007).

In1997, Couser has used form factor to estimate the total viscous component,
where form factor has influenced on frictional resistance of the velocity
augmentation and changes in the boundary layer due to the modified pressure field
around the demihulls (1997). Takinaci et al, has presented to predicted predict the
flow around a three-dimensional rectangular foil section including the effect of
boundary layer used potential-based Surface Panel Method (SPM), where that
method based on boundary-integral formulation (Morino formulation) (2002).

From 1964 until 2007, Savitsky continues to develop a formula to investigate


the characteristics of the planing hull, later this formula is known as Savitsky Method
(SM) (1964, 1985, and 2007).
16

2.4 Longitudinal static stability

The longitudinal static stability characteristics of a Wing-In-Ground (WIG)


effect craft are rather different with the regular airplane .Chun and Chang (2002)
have tried to investigate longitudinal static stability 20-passenger WIG based on
wind tunnel test data, where the sea surface variation is neglected, and these a
surface is treated as a rigid wall. These stability characteristics play an important role
in designing a safe and efficient WIG due to its potential danger in sea surface
proximity.

V.I Korolyoc (1998) has investigated longitudinal stability of ekranoplan and


hydrofoil ships, where all parameter investigation is fulfilled and recommendations
are obtained for the choice of combinations of the carrying system basic elements.

S.C. Rhodes and A.T. Sayers (2009) was investigated the effect of flat
ground on the aerodynamic force slender un-cambered DHMTU rectangular wing
with aspect ratio 3. The investigation were performed with comparison between
experimental work result and Computational Fluid Dynamics (CFD) for investigate
influence of non-flat ground on the performance piece of rectangular wings in
ground condition, and was investigated in an open jet wind tunnel test with a flat
moving ground plane. The position ACH was found downstream of position the
ACP. The SSM were primarily unenthusiastic at all ground clearances (h/c), from
that result that rectangular wing was unstable in ground condition.

Hiromichi Akimoto, Syozo Kubo and Motoki Tanaka (2004), has tried to
applied canard stabilizer instead of horizontal tail and forward mounted propeller in
Wing in Surface effect Ship (WISES) to get stability requirement in the surface
effect.
17

Collu et al (2009) has been tried to solve advance mathematical framework


for the longitudinal static stability of a high-speed craft with planing hull and
aerodynamic surfaces.

2.5 Longitudinal dynamic stability

The dynamics in ground effect is quite dissimilar with the regular airplane
cause influenced by the ground, since the aerodynamic force and moment
coefficients, which in turn depend on height and angle of attack. Rozhdestvensky
(1998) has described longitudinal motion of ekranoplan in extreme ground effect for
very small relative ground clearances using singular asymptotic method, where
derivation of quartic and quintic characteristic equations of unsteady perturbed
motion is offered. In this analysis he shown in very close propinquity to the ground
parameters of stability and motion of the lifting system depend on curvature of the
lower side of the wing to relative ground clearance and ratios of design pitch angle.
In the other hand, the distances of the order of the chord from the moment of
perturbation the vehicle performs matching induced motions in height and pitch in
the same velocity.

The problems of aerodynamics, hydrodynamics and dynamics of ekranoplan


with particular attention to the matters related to use of power augmentation and
peculiarities of static and dynamic stability has been described by Rozhdestvensky
(1998). Nikolai Kornev (2003) has tried to define the mathematical modeling of WIG
motion in all condition, such as planing, take-off, transition to flight, and flight itself.
The theory developed includes nonlinear hydro-aerodynamics base for simulation of
motion here enhances the process of designing WIG vehicles and can be used for
studying emergency problems in WIG operation. The results of numerical have been
18

compared with experimental data obtained for planing and flight condition of
motion.

The dynamics in ground effect has been investigated by Nicola de Divitiis


(2009), where an analytical formulation is proposed for the force and moment
calculation in the existence of the ground and taking the aircraft attitude and sink rate
into account. This feature, the aerodynamic coefficients are firstly calculated for a
representative vehicle and its characteristics in ground effect and performance and
stability characteristics and non-linear dynamics are discussed with reference to
significant equilibrium conditions. Collu et al (2009) has been tried to solve advance
mathematical framework for the longitudinal dynamic stability of a high-speed craft
with planing hull and aerodynamic surfaces. The dynamic stability analyses of a 20-
passenger WIG during cruise has investigated by H.H. Chun, and C.H. Chang
(2002), where the quality of the 20-passenger WIG has analyzed according to the
military regulations.

The basic concept dynamic motion WIG come from conventional dynamic
motion airplane (see Figure 2.5), where mathematical dynamic motion airplane can
be expressed as (Chun 2002, Tulapurkara 2006).

(i)
19

Figure 2.5: Dynamic airplane component (Tulapurkara 2008)

Illustration equilibrium of WIG during take-off can be expressed at Figure


2.6 Kumar (1972), Irodov (1970), Staufenbiel (1980), have been successfully to
introduce height derivatives for the wig in the above equation (Chun 2002).

X (ii)
Xh =
h

Z
Zh = (iii)
h

M
Mh = (iv)
h
20

Figure 2.6: Dynamic WIG Component (2002)

After arranging equation (ii), 0, (iv) into equation (i), mathematical model
dynamic motion WIG craft can be expressed as (Chun 2002):

(v)
CHAPTER 3

RESEARCH METHODOLOGY

3.1 Introduction

In this chapter is describing the research methodology of longitudinal static


stability and longitudinal dynamic stability of a trimaran WIG during takeoff. The
research methodology consists of five main stages to complete the research. The
stages of research are: i) Initial design Model, ii) Computational calculation, iii)
Experimental work, iv) Comparison and analysis of computational results with
experiments or other published researchers, v) Longitudinal static and dynamic
stability analysis.

3.2 Initial design of model

The model of WIG catamaran vehicle of current study designed use Maxsurf,
initial hydrodynamic calculation using Hidromax Maxsurf, for estimating weight
distribution using Workshop Maxsurf. The initial design of model and its principal
dimension are shown in Table 3.1 and Figure 3.1.
22

Table 3.1: Principal dimension of WIG Trimaran

Main Principal Dimension

Length over all (LOA) 1.205 m

Breadth over all (BOA) 0.833 m

Hull breadth (B) 0.133 m

Ram Wing 0.4166 m


Wing Span (b)
Dihedral Wing 0.208 m x2

Ram Wing 0.666 m


Chord length (c)
Dihedral Wing 0.666 m

Figure 3.1: Trimaran WIG design


23

3.3 Computational calculation

Aerodynamic characteristics can be solved by many methods. As mentioned


in the literature many researchers have tried to do it with a variety of methods, such
as the potential flow, particle image velocimetry combining between (PIV) and laser
Doppler anemometry (LDA), the Reynolds Average Navier-Stokes (RANS)
equations, and the Vortex Lattice Method (VLM).

3.3.1 Vortex Lattice Method (VLM)

In this thesis, using a VLM that will be combined with ground effect
Wieselsberger theory and Hemke theory used to determine the effect endplate to
aerodynamic characteristics. Richard and. Lamar method used to analyze the
complex wing shapes. Vortex lattice method was developed in three main category
algorithm using language programming Matlab to make certain code portability
transversely model, where categories are the preprocessor, the solver and the
postprocessor.

3.3.1.1 Preprocessor Vortex lattice Method (VLM)

The preprocessor have function to set up the vertices and the boundary
environment beginning inputs. The preprocessor has several steps; such as input,
layout and meshing. See Figure 3.2 for illustration.
24

Figure 3.2: Flowchart preprocessor

3.3.1.2 Solver Vortex lattice Method (VLM)

The solver has function to convert the conciliator consequences of the


preprocessor (meshing) into the aerodynamic force, where all information from the
preprocessor was described all essential data for the arithmetical approach. The
solver has several steps, such as: downwash, boundary condition, inwash &
sidewash, force and coefficients computation. See Figure 3.3 for illustration.

Figure 3.3: Flowchart processor


25

3.3.1.3 Postprocessor Vortex lattice Method (VLM)

The postprocessor's responsibility is to show the computed outcome in an


inclusive way. In the other hand there are some problem to categorize data and to
select the best display of plot, expected that all problems can be solved in this phase.
This of course depends on the language used. See Figure 3.4 to illustrate flow of
postprocessor.

Figure 3.4: Flowchart postprocessor

3.3.2 Planing hull

The calculation planning hull using the Savitsky method was set up in
Microsoft excel combine with Microsoft Visual Basic for iteration. The develop code
for this program similar with vortex lattice method, with three main phase,
preprocessor, processor and postprocessor. See Figure 3.5 for flowchart illustration.
26

Figure 3.5 Flowchart Savitsky method

3.4 Experimental work

Experimental Work will be do using the wind tunnel test LST-UTM to get the
aerodynamic force and the free running test to get the hydrodynamic force Lift force,
drag force, moment and centre of pressure of wing of a Trimaran WIG model (foil
NACA 6409) are experimentally will be done in LST-UTM wind tunnel with wind
speed 25.5 m/s, where its generated by a single stage fan that delivering maximum
air speed 80 m/s and propeller by 430 KW AC motor from ABB with Allen Bradley
system. Aerodynamic characteristic are deliberate with various primary aerodynamic
parameters such as clearance of ground, angle of attack and aspect ratio (AR) in test
section 2.0 m wide x 1.5 m high x 5.5 length with windows on both sides, settling
chamber with a acceptable flow quality, where velocity uniformity <0.15,
temperature uniformity <0.2C, flow angle uniformity <0.15 and turbulence
intensity <0.06%). Visualizing the flow of the air including the tip vortices around
the wing and wake subsequent the wing during test has been done using smoke trace
test and the model test has support with 3-strut support system with various ground
clearance (h/c) and angle of attack (). Aerodynamic forces, moments, and pressure
distribution are measured with electronic pressure scanner 128 port scan valve
27

system and 6-component balance with accuracy of the balance within 0.04% based
on 1 standard deviation. All data test have been read and stored at control room with
fully integrated automatic control.

Hydrodynamics drag obtained from the free running test. Telemetry system
used to obtain the data directly. Then the data were found will be converted using
equation from the calibration rig.

3.5 Comparison and analysis of results

In this stage of comparison of the computational results and experiments or


published work, analysis and discussion and then writing report. According of
experimental results will be tried to find the best computational simulation.

3.6 Longitudinal static stability & longitudinal dynamic stability analysis

The last stage is presenting integrated numerical equation to study


longitudinal static stability and longitudinal dynamic stability WIG trimaran with
planing hull. In this stage static stability margin was modified with new
configuration where, planing hull was added into that configuration. Then, new
configuration dynamic motion during take-off was created after undertaking the
classical WIG motion by adding hydrodynamic characteristic, so, the complete
equations motion of a trimaran WIG during take-off solved and investigated
28

GET DATA
GET DATA
(NUMERICAL)
(EXPERIMENT)

VALIDATE DATA
N VALIDATE DATA
(PUBLISH WORK & N
(MEASUREMENT SET
COMMERCIAL
UP AND FLOW

Y
Y

AERODYNAMIC & HYDRODYNAMIC


AERODYNAMIC & HYDRODYNAMIC

RESULT
RESULT

VALIDATION

N CHECK N
DATA

NEW CONFIG.SSM ANALYSIS LONGITUDINAL CHECK ROUTH HURWITZH


STATIC STABILITY CRITERIA

ANALYSIS LONGITUDINAL CHECK CAP CCRITERIA


NEW CONFIG.DM
DYNAMIC STABILITY

SUMMARY & DISSCUSION

Figure 3.6: Flowchart research


CHAPTER 4

MATHEMATICAL MODEL DEVELOPMENT

4.1 Introduction

This chapter presents the formulation for the aerodynamic characteristic,


hydrodynamic characteristic, longitudinal static stability and longitudinal dynamic
stability WIG during take-off.

4.2 Vortex Lattice Method

The traditional representation for flat wing is shown in equation (vi), see
Figure 4.1 for illustration (Margason 1971):

w cos U sin = 0 (vi)


30

Figure 4.1: Flat Wing Sketch (Margason 1971)

The boundary condition can be extended to represent wing with dihedral in


equation (vii), see Figure 4.2 for illustration.

w cos cos v sin U sin cos = 0 (vii)

Figure 4.2: Flat Wing with Dihedral Sketch (Margason 1971)


31

Equation (viii) show wing from upstream toward the trailing edge of the right
half of the wing span and see Figure 4.3 for illustration.

w cos cos r v sin r U sin cos r = 0 (viii)

Figure 4.3: Looking from upstream toward the trailing edge (Margason 1971)

Convention geometry can be expressed as:

= l = r (ix)

For small angle of attack

W = v tan U (x)
32

The vortices lattice is located in plane parallel to free stream as show in


Figure 4.4.

Figure 4.4: Vortex filaments (Margason 1971)

Related from Figure 4.4, equation (viii) (ix) (x) can be expressed for
downwash influenced coefficient:

Fw ( x ' , y , z , s , , ' ) = (xi)

( y tan ' x ' ) cos


( x ) + ( y sin ) + cos ( y tan 2 + z 2 sec 2 ) ' 2 z cos sin ( y + x ' tan ' )
' 2 2 2 2

x
( )
x ' + s cos tan ' cos tan ' ( y + s cos ) cos + ( z + s sin )

[(
x ' + s cos tan '
2 2
)
+ ( y + s cos ) + ( z + s sin )
2
]

x

( '
s cos tan '
)cos tan '
(y s cos )cos + (z s sin )

[(x '
s cos tan '
) 2
+ (y s cos
2
) + (z s sin )2
]
33

( y s cos ) x ' s cos tan '



2
[( y s cos ) + (z s sin ) ] (x ' s cos tan ' )2 + ( y s cos )2 + (z s sin )2
2
[ ] 1
2

( y + s cos ) x ' + s cos tan '



2
+
[( y + s cos ) + (z + s sin ) ] (x ' + s cos tan ' )2 + ( y + s cos )2 + (z + s sin )2
2
[ ]
1
2

For sidewash influenced coefficient can be written as;

Fv ( x ' , y , z , s , , ' ) = (xii)

x ' sin z cos tan '


( x ' ) 2 + ( y sin ) 2 + cos 2 ( y 2 tan 2 + z 2 sec 2 2 yx ' tan ' ) ' 2 z cos sin ( y + x ' tan ' )

x
x
( '
+ s cos tan '
)cos tan '
(y + s cos ) cos + (z + s sin )

[(x '
+ s cos tan '
) 2
+ (y + s cos
2
) + (z + s sin
2
) ]



(x '
s cos tan '
)cos tan '
(y s cos ) cos
+ (z s sin )

[(x '
s cos tan '
)
2
+ (y s cos
2
) + (z s sin 2
) ]

( y + s cos ) x ' + s cos tan '



+
[( y + s cos )2 + (z + s sin )2 ] [(x '
+ s cos tan '
)2 2
+ ( y + s cos ) + ( z + s sin )
2
] 1
2

( y s cos ) x ' s cos tan '



2

[( y s cos ) + (z s sin ) ]
2
[(x '
s cos tan ' 2
) + ( y s cos ) 2
+ ( z s sin )
2
]
1
2

For backwash influenced coefficient can be written as;

Fv ( x ' , y , z , s , , ' ) = (xiii)


z cos y sin
( x ' ) 2 + ( y sin ) 2 + cos 2 ( y 2 tan 2 + z 2 sec 2 2 yx ' tan ' ) ' 2 z cos sin ( y + x ' tan ' )
34


x
(x '
+ s cos tan '
)cos tan '
(y + s cos ) cos + (z + s sin )sin

[(x '
+ s cos tan ' ) +
2
(y + s cos )2 + ( z + s sin ) 1
2
]



(x '
s cos tan '
)cos tan '
(y s cos ) cos + (z s sin ) sin

[(x '
s cos tan ' ) +
2
(y s cos )2 + ( z s sin ) 1
2
]

x
(
x ' + s cos tan ' cos tan ) '
(y + s cos ) cos + (z + s sin )sin

[(
x ' + s cos tan '
2
+ ) (y + s cos
2
) + (z + s sin )
1
2
]



(x '
s cos tan '
)cos tan '
(y s cos ) cos + (z s sin )sin

[(x '
s cos tan ' ) +
2
(y s cos )2 + ( z s sin ) 1
2
]

We have known, infinitive vortex lines induces a flow field have been
expressed by Biot-Savart, and have been modified by Bertin & Smith. It takes the
form of equation:

1 r .r r1 r2 (xiv)
U= n 1 2 r0
4 [r1 .r2 ]2 r1 r2

From equation (xi) (xii) (xiii) (xiv), downwash, sidewash, backwash velocity
could be expressed as:

1 (xv)
U u ,v , w = n Fu , v , w ( x , y , z , s , , )
4

Using equation (x) (xv)can be written as:


35

n (xvi)
(Fw Fv tan ) = U
4

For a vortex lattice of N element and symmetrical aerodynamic loading on


each half of the wing, equation (xvi) can be expressed as:

N
2

n =1 4
(Fw ,n Fv ,n tan n ) = U (xvii)

The lift per unit length of a vortex filament for wing with dihedral can be
written as Kutta-Joukowski theorem (Margason 1971):

l ~ = V (xviii)

The coefficient along an elemental length of chord wise and coefficient


divided by free stream dynamic pressure can be expressed to implement equation
(xviii) as:

lt~ 2 V (xix)
Clt = = cc
qSref Sref U U

Like equation (xix) the expression a spanwise bound vortex filament divided
by free stream dynamic pressure and reference area can be expressed as:

l s~ 2 u v
Cl s = = (2 s ) 1 + tan cos (xx)
qSref Sref U U U
36

Finally, the total lift coefficient in out ground effect can be expressed as:

N /2
2 l s~ lt~
Cl (OGE ) = 2 +
n S qSref n qSref n (xxi)

4.3 Effect of flat ground and endplate on aerodynamic characteristic

Regarding literature review effect of flat ground will be explain base on


Wieselsberger theory and effect endplate base on Paul E. Hemke Theory. Detail
explanation will be present in sub chapter below.

4.3.1 Effect flat ground

The theoretical treatment of ground effect presented by Wieselsberger [0] for


predicting the reduction in induced drag for wing at various heights of the quarter
chord of the wing above the ground . Experiments have been full field by Elliot
G.Reid (1937) and Arthur W Carter (1961). The induced drag in ground effect is
given by the equation:

2
C S (xxii)
C Di ( IGE ) = (1 ) L 2
b
Where:
37

h
1 1.32
= c (xxiii)
h
1.05 + 7.4
c

Equation (xxii) can be expressed as:

b2 1
C L ( IGE ) = C Di (xxiv)
s 1

Related equation (xxi)and (xxiv) coefficient in ground effect could be written


as:

1
C L ( IGE ) = C L ( OGE ) (xxv)
1

4.3.2 Effect endplate

The frictional drag of the endplates can reduce induced drag, is sufficiently
large to increase the efficiency of the wing. Analysis of experimental drag of wing
with endplate has done by Paul E. Hemke (1927).

2
C S
C Di (WE ) = (1 ) L 2 (xxvi)
b

Where:
38

2he (xxvii)
1.66
= c
2h
1 + 1.66 e
c
Just like equation (xxv) coefficient with endplate could be shown:

1 (xxviii)
C L (WE ) = C L ( NWE )
1

4.4 Step by step Savitsky method

The elemental hydrodynamic characteristics of prismatic planing hulls and


the empirical planing equations involved are thoroughly described in Savitsky. The
equations given, describe the lift, drag, wetted area, centre of pressure and propoising
stability limits of the planing hull, as a function of speed, trim angle, deadrise angle
and loading. These empirical planing equations are combined so as to formulate a
computational procedure in the form of a table where forces and moments acting on
a planing hull are considered, in order to determine an equilibrium trim are
applicable only to the bottom pressure area aft of the leading edge stagnation line. On
that basis a prediction can be made on resistance, effective power, running trim, and
draft and porpoising stability of a prismatic planing hull. His calculations show the
following:

i. For the first time we should know centre of pressure planing hull. Centre of
pressure for a flat planing surface is given by J.B. Hadler (1966):
39
LCP 1
Cp = 0.75 ( ) (xxix)
LM Cv 2
5.21 + + 2.39
2

Where:

( LK + LC )
= (xxx)
2B

b tan (xxxi)
L K = LM +
2 tan

b tan (xxxii)
LC = LM
2 tan

The position center of pressure can be predicted:

LCP = C P = C P LM
(xxxiii)

Planing hull also has hydrodynamic lift; Savitsky method involves the
following empirical equations:

For speed coefficient or Breadth Froude Number is given by the


following equation D. Savitsky (1927):

Vs
Cv = (xxxiv)
gb
40

The lift coefficient for flat planing hull (zero dead rise) is given by the
following equation Savitsky (1964):

Displ
C Lo = (xxxv)
0.5 V 2 b 2

Also can be described with parameter angle of trim and speed coefficient by
Savitsky (1964):

0.00552.5
C Lo = 1.1 (0.0120.5 + 2
) (xxxvi)
Cv

For Vee-surface (prismatic), the lift Coefficient has developed by D.


Savitsky [0] as:

0.6
C L = C Lo 0.0065 C Lo (xxxvii)

Then calculate the magnitude of average bottom velocity VM for a planing


surface is simply developed for a given vessel speed (VS) (Hadler 1966):

0.5
VM 0.0120.5 1.1 0.0065 (0.0120.5 1.1 ) 0.6 (xxxviii)
1
VS cos

The total hydrodynamic drag of a planing surface is composed of pressure


drag developed by pressure acting normal to the inclined bottom and
41

viscous drag acting tangential to the bottom in both of pressure area and
spray area Savitsky (1964).

Df
D = tan + (xxxix)
cos

C f V12 (b 2 )
Df = (xl)
2 cos 4

Where Cf is applied according ITTC, 1959 friction line, and is given by the
following:

0.075
Cf = (xli)
( LogRe 2) 2

The Reynoldss Number can be solved by applying the formula:

VM xLM xbxVs V
Re = = x M (xlii)
VS

We must calculate equilibrium at a reasonable trim angle, the formulas has


developed by J.B Hadler (1966).

Calculate Total Moment from moment due to displacement and moment due
to drag force:
42

M T = M + M Df (xliii)

Where:

c cos( + ) f sin (xliv)


M =
cos cos

f
(xlv)
M Df = Df a c tan
cos

c = LCG LCP (xlvi)

Total Resistance at equilibrium trim angle can be calculated using:

cos( 0 + )
RT =
cos
[
sin 0 + D f 0 ] (xlvii)

Where:

M BH 1 ( 2 1 ) (xlviii)
0 = 1
M BH 2 M BH 1

Df 2 Df1
Df 0 = Df1 + 0 1 (xlix)
2 1

Thrust can be calculated using Fox & Blound Formula (1976):


43

TT = RT xMfactor (l)

Where:

1 LSP LSP 3( Fn 0.85)


1.45

Mfactor = 0.98 + 2 e 2( Fn 0.85) 3 e (li)


2 B B

4.5 Longitudinal static stability

WIG should be stable in pitch like an airplane, if rigid body have been
disturbed should be return to undisturbed position (Chun 2008), and mathematically
give as:

C M < 0 (lii)

dCm dCm dCl


CM = = x <0 (liii)
d dCl d

And derivatives could be written as:

dCM
CM = CL < 0 (liv)
dCL

If:
44

C L > 0 (lv)

So:

dC M (lvi)
<0
dC L

Where:

dC M X X ac (lvii)
= cg
dC L c

And derivatives could be written as:

dC M X cg X ac (lviii)
= <0
dC L c

From equation (lviii) it can be deducted the position of center of gravity


(COG) of the craft in front of the aerodynamic center of pitch (ACP).

WIG also need to have stability in the vertical dimension. The mathematical
can be expressed as Kumar (1968):

CMh < 0 (lix)

dCm dC m dCl (lx)


C Mh = = x <0
dh dCl dh
45

And derivatives could be written as:

C M
C Mh = C Lh < 0 (lxi)
C L

The derivatives of the lift coefficient with increasing height must be


negative, mathematical have been expressed by Irodov (1970) and Staufenbiel
(1980).

C Lh < 0 (lxii)

So:

dC M (lxiii)
>0
dC L

Where:

dC M X Xh (lxiv)
= cg
dC L c

Refer to equation (lviii) derivatives could be written as:

C M X cg X h (lxv)
= >0
C L c
46

From equation (lxv) we can describe the position center of gravity (COG)
craft behind of aerodynamic center in height (ACH).

After rearranging equation (lviii) to equation (lxv), height stability equation


can be derived as:

(lxvi)
X ac X h > 0

The distance between the aerodynamic center in pitch and aerodynamic


center in height are referred to as Static Stability Margin (SSM) (Rhodes 2009)

(lxvii)
SSM = X ac X h > 0

The above formula only for cruise condition but for takeoff condition,
hydrodynamic forces also affect the static stability margin. Planing hull have pitch
stability, and mathematically give as (Savander 2003):

CMH < 0 (lxviii)

And derivatives could be written as:

C M
C MH = CLH < 0 (lxix)
CL
47

If:

CLH > 0 (lxx)

So:

dC M (lxxi)
<0
dC L

Where:

C M X cg X H (lxxii)
=
C L c

Refer to equation (lviii) and equation (lxv) derivatives could be written as:

C M X cg X H (lxxiii)
= <0
C L c

From equation (lxxiii) we can describe the position center of gravity (COG)
of the craft should be located upstream of the Hydrodynamic center of pitch (HCP).

The position HCP WIG during take-off similar with prediction Collu et Al [0]
for AAMV, where the position hydrodynamic centre in heave (HCH) should be
located downstream of the aerodynamic centre in height (ACH) and aerodynamic
centre of pitch (ACP).
48

According both statements we can conclude the position aerodynamic centre


in height (ACH) should be located upstream of the centre of gravity (COG) of the
craft , aerodynamic centre of pitch (ACP) & hydrodynamic centre in Pitch (HCP)
should be located downstream of centre of gravity (COG) of the craft. So, new
configuration Static Stability Margin (SSM) a trimaran WIG during take-off can be
written as:

SSM = X ac X h + X H > 0 (lxxiv)

Positive and negative signs from above formula, only to indicate the position
of force in front of or behind the centre of gravity (COG), positive for behind COG,
negative for in front of COG.

The new configuration static stability margin can be illustrated from the
Figure 4.5.

Figure 4.5: New Configuration SSM during take-off


49

4.6 Longitudinal dynamic stability & dynamic motion

Configuration used to create mathematical models of motion on the WIG


during the take-off is very different with cruising. Because at during take-off not
only the aerodynamic force that influence but also influence the hydrodynamic force.
Conditions are very similar to aerodynamically alleviated marine vehicle (AAMV),
where mathematical models have been created by M.Collu et.al (2009). The factors
that influence in this analysis are aerodynamic force, hydrodynamic Force,
propulsion system, control system, disturbance, and gravity. Total Force can be
illustrated as:

F = Fa + Fh + Fp + Fc + Fg + Fd (lxxv)

Each force with equilibrium state with subscript (0) and perturbation from the
datum with subscript () can be expressed as;

F = F0 + F ' (lxxvi)

4.6.1 Aerodynamic force

The traditional aircraft motion can be expressed like equation (i) and after
add height derivatives into that equation, equation WIG motion become equation (v),
after taking into equation (lxxvi) and Tailor linear expansion equation can be
expressed as:
50

F a = F0a + F a ' (lxxvii)

Where:

F0a = ( X 0a Z 0a M 0a ) T (lxxviii)

a a a
'
a a a
'
X . X . X . . X u.. X w.. X q.. ..
u u u
w q
X ha
a ' a . a ..
F a' = Z h h + Z . Z a. Z a. a a
w + Z u.. Z w.. Z q.. w (lxxix)
M a u w q
. .
h a
q q..
M . M a. M a. M ..a M a.. M ..a
u w q
u w q

4.6.2 Hydrodynamic force

Collu et al (2009) was proposed planing craft hydrodynamic force with


added mass and damping coefficients with respect to the aero-hydrodynamic
axis system, where that forces are extended in a Taylor series through the third
order equations of motion and was written as discrepancy equations with steady
coefficients with linearize the non-linear scheme of equations of motion and
derivatives are separated in restoring coefficients.. That equation expressed as:

(lxxx)
F h = F0h + F h '
51

Where:

F0h = ( X 0h Z 0h M 0h ) T (lxxxi)

h h h
'
h h h
'
h h
0 X w X q X . X . X . . X u.. X w.. X q.. ..
u u u
' u w q

h' h h
F = 0 Zw Zq w + Z h Z h Z h
. h h h
..
u. . . w + Zu.. Zw.. Zq.. w (lxxxii)
q w q
. ..
0 M h M h h q h q
w q
M . M h. M h. M .. M h.. M ..h
u w q
u w q

4.6.3 Propulsion system, control system and disturbance

In this thesis thrust from propulsion system equal with total drag WIG during
take-off. Control force is zero with assume control system are fixed. Disturbance
from environmental, and other disturbance are neglected. Forces are expressed as:

F p = F0p (lxxxiii)

F c = F0c (lxxxiv)

Fd = 0 (lxxxv)
52

4.6.4 Gravitational force

Contribution force from gravitational force can be expressed as:

F g = F0g + F g ' (lxxxvi)

Where:

F0g = (0 mg 0) T (lxxxvii)

F g ' = ( mg 0 0 ) (lxxxviii)
T

4.6.5 Equilibrium state, linearized, and cauchy standard equation of motion

Equilibrium state equation motion a trimaran WIG during take-off can be


reached after under taking equation (lxxix), (lxxxii), (lxxxiii), (lxxxiv), (lxxxv),
(lxxxvi) into longitudinal linearized equations of motion with accelerations are 0 and
all the perturbations velocities and the perturbation characteristic (RULM standard).
So equation can be expressed as:
53

0 = X 0a + X 0h + X 0p + X 0c + X 0d + X 0g
0 = Z 0a + Z 0h + Z 0p + Z 0c + Z 0d + Z 0g
(lxxxix)
0 = M 0a + M 0h + M 0p + M 0c + M 0d + M 0g

Where longitudinal linearized equations of motion (assume system equation


is decoupled) can be expressed as (superscript .representing the perturbated state
will be omitted):

..
X = m u'
.. .
Z = m( w' q 'V0 )
..
M = I q q' (xc)

After under taking equation (xc) into equation (lxxxix) and eliminating the
insignificant conditions, the numerical equation of motion a trimaran WIG during
take-off written in the aero-hydrodynamic axis system can be shown as:

.. .
u u u
.. . w + [D ]h = 0 (xci)
[A] w + [B ] w + [C ]
.. .
q q q

Where, matrix [A] is the amount of the mass matrix, the hydrodynamic added
mass derivatives and the aerodynamic added mass. Usually in aerodynamics
aerodynamic added mass they are not called added mass terms, but simply
acceleration derivatives, so matrix [A] can be expressed as:
54

'
h a h h

m X .. X .. X .. X ..
u w w q

(xcii)
A = Z ..h m Z a.. Z h.. Z ..h
u w w q


M ..h M a.. M h.. I 55 q..
M h

u w w

Matrix [B] defines as the damping matrix, equation expressed as:

'
a h a h a
h
X . X . X. X. X. X.
u u w w q q


B = Z a. Z h. Z a. Z ..h a
Z. Z. h
(xciii)
u u w w q q


M .a M .h M a. M h. M a. M h.
u u w w q q

Matrix [C] defines as the restoring matrix, equation expressed as:

'
h
h
0 X w X q

h
C = 0 Z wh Zq (xciv)

0 M wh h
Mq

Matrix [D] defines as effect of the ground clearances above the surface on the
aerodynamic characteristic, equation expressed as;
55

'
h h

0 X w X q

h (xcv)
C = 0 Z wh Zq

0 M wh M qh

The structure of numerical equations of motion a trimaran WIG during take-


off in cauchy standard expressed as:

'

[A] [B] [A] [C] [A] [D]
1 1 1


(xcvi)
H = [I ]3x3 [0]3x3 [I ]3x1

0 0 1 0 0 V0 0

All derivatives can be seen at Appendix C.


CHAPTER 5

EXPERIMENTAL WORK

5.1 Wind Tunnel Test

The wind tunnel is tunnel where model of an airplane or object test set in
there and windblown through it at a certain speed to study the way air moves around
the model. The wind tunnel has function to measure air velocity and pressures model
of an airplane or object test, where air rapidity during the investigation (called the
throat) is dogged by Bernoulli's principle, the dynamic pressure, the static pressure,
and the temperature measure in the air stream for compressible flow only. The route
of air stream roughly a model test can be resolute by tufts of thread attached to the
aerodynamic model surfaces and aerodynamic model surface can be visualized by
increasing air stream around test model, also can be use smoke trace or bubbles trace
to visualized air stream on test model. Pressures distributions on the test model are
regularly considered with beam balances where the model test connect to strut,
strings, and multi-tube manometers. Also can be considered using pressure-sensitive
paint where elevated restricted force is indicated by lowered fluorescence of the
paint at that point. Sensitive pressure belts furthermore can be utilize for determine
pressure distribution and multiple ultra-miniaturized pressure sensor modules are
incorporated into a stretchy shred, where the shred is close to the aerodynamic model
surface with adhesive and it transfer data depicting the distribution of pressure along
its model surface. There are various techniques can be used to study the actual air
57

stream roughly model test and system of process using Reynolds number and Mach
number (2008).

5.1.1 Facility Low Speed Wind Tunnel Universiti Technology Malaysia (LST -
UTM)

Low Speed Wind Tunnel University Technology Malaysia (LST -UTM) has
capability to deliver airspeed inside the test section with maximum speed 80 m/s,
where test section size area is 2.0 m wide x 1.5 m height x 5.5 m length. LST -UTM
have high accuracy and good repeatability of wind tunnel test results with capability
an excellent flow quality, where flow uniformity less than 0.15%,temperature
uniformity less than 0.2, flow angularity uniformity less than 0.15, and turbulence
less than 0.06%). LST-UTM also has capability to provide wide range of testing
include aircraft, ground surface vehicle and industrial aerodynamics such as building,
bridges, street-lantern light and wind turbine. LST-UTM has three important
components in the wind tunnel which is known as test section, fan-motor and settling
chamber (see Figure 5.2) (2008).

For detail LST-UTM see specification below:

Date Built : 2001

Size H/W/L : 4.9 ft x 6.6 ft x 19.0 ft (1.5 m x 2 m x 5.8 m)

Max Wind Speed : 3 to 80 m/s (Mach 0.23)

Construction Type : Solid


58

Reynolds Number : 1,000,000

Circuit Type : Closed circuit, single return

Power : 430 kW AC motor

Power Type : 2.9 m diameter, 16-blade, axial fan

Pressure/Temperature : Atmospheric

Data System : PC-Dell, Windows 2000/NT Servers

Data Acquisition : Pacific Instrument PI 6000 Series

Pressure System : Scanivalve Model DSM3000 ZOC Pressure


transducer (128-port)

Balances : 6-component external, internal, semi-span

Ground Effects : Not Available

Test Types : Aero-tests (3D, 2D, half-model), Ground


vehicle (automotive)

Air Supply System : 12 bar

LST-UTM is shelter inside the Aeronautical Engineering Laboratory building


UTM and furnished with compress air facility for general purpose applications. The
test section is connected to wind tunnel control room via metal structure platform
(Figure 5.1) (Suhaimi 2008).
59

Figure 5.1: Layout LST-UTM (Suhaimi 2008)

Figure 5.2: LST-UTM main component (Suhaimi 2008)


60

5.1.2 Procedure wind tunnel test at LST-UTM

Lift force from airfoil generally developed by lower pressures above the wing
(downwash) below the wing (upwash) and side the wing (sidewash) with deference
to the pressure of the resembling air. Lift coefficient increases with higher angle of
attack up to an essential angle and ground clearance. This essential angle the lift
coefficient decreases drastically, generally are called stalled. Generally pressure
distribution considered with little tubes implanted in model test and connected to
anxiety transducer with 128 inlet port, where inlet port are placed at various percent
chord on both superior and inferior surfaces. The distribution of pressure ought to
extrapolate the previous dimensions on the superior and inferior surfaces of the
model test, where coefficient of pressure (Cp=0) at the trailing edge of the model
test. Function of the impetus theory will be use to indicated drag coefficient on the
model test, where air stream ought to equivalent with lessening in linear impetus that
flow stream. Since the air stream resembling the model test is homogeneous, the
coefficient drag could be written as (Rae 1984):

Yw
2
YW 1 (xcvii)
CD =
c qc qdy
Yw

2

Yw = distance across of wake, q = restricted dynamic force, q = liberated


flow restricted dynamic force.

The wake assessment is through as far from behind the model test as probable
to deliver the static pressure possessions insignificant. On the testing could be verdict
that the wake turf the coefficient drag obtained in the held up arrangement disagrees
with additional drag, and they ought to observe the assumptions completed in the
61

impetus equilibrium investigation in order to recognize which of these assumptions


are dishonored by the stream above a stalled model test (Rae 1984).

1 P
C mLE =
c2 q x(dc) (xcviii)

x = spacing from in front of wing.

5.1.2.1 Preparation model

Model for wind tunnel test should have significant strength to suffer
aerodynamic forces. The opportunities of the material selection to make the model
are aluminum with scale 1: 10, for detail specification model test see Table 5.1.
Model for wind tunnel test made use aluminum block ( Figure 5.3), than shaping
with 3 axis CNC Milling MAHO 500 E2 (Figure 5.4) with cold milling technique
(Figure 5.5). Finally model can be show at Figure 5.6.

Table 5.1: Model for wind tunnel test

Main Sizing Dimension

Ram Wing 0.41666 m


Wing Span (b)
Dihedral Wing 0.20833 m x2

Ram Wing 0.66667 m


Chord length (c)
Dihedral Wing 0.66667 m
62

Figure 5.3: Aluminum block for model wind tunnel test

Figure 5.4: CNC milling MAHO 500 E2


63

Figure 5.5: Cold milling

Figure 5.6: Wing model for wind tunnel test


64

5.1.2.2 Set up Instrumentation

ii. Set up fan motor and drive system


First time we must set up fan motor and drive system to get wind speed
on our experiment, where LST-UTM is propelled by ABB motor with
maximum power 430 kW AC and control with Allen Bradley drive
system. Power consumption of the motor with test section wind speed is
show in Figure 5.7 (Suhaimi 2008).

Figure 5.7: Set up fan power for wind speed (Suhaimi 2008)

iii. Set up chamber


Settling chamber is an area where the flow is control using honeycomb to
straighten the flow and three screens to reduce the turbulence intensity
(Figure 5.8). Heat exchanger was used to reject heat generates from air
friction and maintain the test section temperature at normal condition
(Figure 5.9) (Suhaimi 2008).
65

Figure 5.8: Honeycomb and screen LST-UTM (Suhaimi 2008)

Figure 5.9: Heat exchanger LST-UTM (Suhaimi 2008)

iv. Set up balance and model support


LST-UTM equipped with a 6-component balance for load measurements.
LST-UTM balance is a pyramidal type with virtual balance moment at
66

the centre of the test section and has a capability to measure aerodynamic
forces and moment in 3-dimensional (Figure 5.10). LST-UTM also can
be tested at various wind direction by rotating the model via turntable
with accuracy of the balance is within 0.04% based on 1 standard
deviation and maximum load range is 1200N for axial and side loads.
LST-UTM balance originally furnished with 3-strut support system (
Figure 5.11). Also provide street lantern light using single-strut support
(Figure 5.12) (Suhaimi 2008).

Figure 5.10: Balance and support system LST-UTM (Suhaimi 2008)


67

Figure 5.11: 3-Struts support (Suhaimi 2008)

Figure 5.12: Street lantern light using single-strut support (Suhaimi 2008)
68

v. Set up control system


LST-UTM equipped with fully integrated automatic Control, data
Acquisition and Reduction System (DARS). The operation and post-
processing data of the tests, measurement of air properties, 6-component
balance, pressure transducer, hot-wire and others transducer can be read
and stored are done inside the control room (see Figure 5.13) (Suhaimi
2008)

Figure 5.13: Data Acquisition and Reduction System LST-UTM (Suhaimi 2008)
69

vi. Set up pressure measurement


Set up surface pressure measurement using electronic pressure scanner
128 port scanivalve system, 2 x 64-port Electronic Pressure
Scanner(Scanivalve DMS 3000 ZOC Series), pressure rating 1 psid, 5
psid FS, +/-0.08 FS accuracy (see Figure 5.14) (Suhaimi 2008).

Figure 5.14: Pressure measurement tools LST-UTM (Suhaimi 2008)

vii. Set up Flow Visualization


Smoke trace is one of a qualitative technique to visualize the air stream
on the model test surface (Figure 5.15) (Suhaimi 2008).

Figure 5.15: Flow visualization LST-UTM


70

5.1.2.3 Testing

The experiment is carried out at the wind speed of 25 5 ms-1. For this class of
experiment, the determining apparatus used is a 3-Components Balance which is
capable of determining aerodynamic force (Suhaimi 2008).

i. Barometer and thermometer have been use in the laboratory to determine the
density of the air stream in section area test.

ii. Pitot probe (Figure 5.16) or air stream check apparatus (Figure 5.17) have been
use to calibrate the section area of wind tunnel by create a graph of velocity
(m/sec) versus frequency motor (0- 60 Hz) and processed that result using
Bernoulli's equation.

Figure 5.16: Pitot probe (Suhaimi 2008)


71

Figure 5.17: Air stream check apparatus (Suhaimi 2008)

1 1
Pa + aU a2 = Pb + bU b2 where (xcix)
2 2

Ua = 0 U b = U (c)
,

1
Pa Pb = U 2 = q or (ci)
2

Pa Pb = 1 hg e and (cii)

1
2 (Pa Pb ) 2
U = or (ciii)

1
hg e 2
(civ)
U = 1

72

iii. Use standard propagation of error analysis to estimate the error in U.

1
f
(x )
i 2
Y =
2
i (cv)
i xi

Y = f ( xi ) (cvi)

iv. A pressure wing is mounted vertically in the wind tunnel. The pressure tubes
(128 with locations indicated above) from wing are connected to the inlet
nipples of the multiplexed tunnel pressure sampling system. The static pressure
of the test section is connected to the reference connection of the pressure
transducer. The dynamic pressure of the air stream is measured with the Pitot
probe. Dividing the pressure measured with the sampling system by gives the
pressure coefficient at the point of the measurement.

v. Operate the tunnel at airspeeds of 25.5 m/sec and make pressure measurements
on the wing at angles of attack of 0, 1, 2, 3, until 10. Check the zero
velocity pressure measurements from the wing and pitot probe before each data
set, measure and correct for any offsets in the pressure transducer at zero
velocity.

5.1.2.4 Gathering data

Collecting data is very important, because the wrong data will cause errors in
the study. There are several stages in the data collection (Suhaimi 2008).
73

i. Plot the pressure coefficient data points (upper and lower surface) as a
function of distance along the chord line of the wing and integrate to find
the Normal Force coefficient Cn which is given as:

c
1
C n = (C PL C PU )dx (cvii)
c0

Find this normal force coefficient for all angles of attack, ground
clearance and wind speeds.

ii. Determine the Lift coefficients CL from CN and Plot CL versus angle of
attack and ground clearance for each wind speed. Show the results on one
graph for comparison purposes.

iii. On a separate graph plot CD versus angle of attack and ground clearance
for each wind speed. Note that this method does not measure drag viscous
forces due to shear stresses and thus may underrepresented the total drag
force on the wing. The drag force and drag coefficient measured in this
experiment is the component of the normal force in the direction parallel
to the free stream flow, and increases as the angle of attack increases.

iv. Using the equation for the pitching moment

c
1
Cmref = 2 (C PL C PU )(X X ref )dx (cviii)
c 0
74

a. Calculate the leading edge pitching moment C mLE for the airfoil

as a function of angle of attack for each velocity and present the


results in a table.

X cp CmLE
= (cix)
c CN

b. Calculate the center of pressure for the airfoil as a function of


angle of attach and ground clearance for each velocity and present
the results in a table.
c. Calculate the pitching moment through the quarter chord point
as a function of angle of attack and ground clearance for each
velocity and present the results in a table.

v. Operate the tunnel at airspeeds of 25.5 m/sec and make pressure


measurements on the wing at angles of attack of 0, 1, 2, 3, until 10.
Check the zero velocity pressure measurements from the wing and pitot probe
before each data set, measure and correct for any offsets in the pressure
transducer at zero velocity.

a. Plot the normalized wake survey pressure distribution versus distance


across the wake for each angle of attack and ground clearance for the
three different velocities.
b. Calculate the drag coefficient for the airfoil by the momentum method
by integrating the wake pressure distribution for each angle of attack and
the three velocities. Plot this drag coefficient versus angle of attack for
each of the flow velocities used in the experiment, and compare this
result to the drag force computed in part 3 above. Make sure to comment
on any differences or discrepancies in the two different drag estimates.
75

vi. Determine the effect of Reynolds Number on lift, drag, and 1/4 cord
pitching moment coefficients. (Plot the lift and pitching moment coefficient
from the pressure wing measurements, and the drag coefficient from the wake
measurements versus Reynolds number for all available angle of attack and
ground clearance.)

5.1.2.5 Correction, normalization, and linearization data

After we get and process result from wind tunnel, results must be corrected
with several corrections. Ishak et al (2006) has been corrected result from LST-UTM
for blockage effect, buoyancy effect, wall interference correction and STI (Strut, Tare
and Interference) correction, because all effect are donate rather considerably to the
finishing result.

i. Blockage Effect
There are two type blockages during experiment, such as: Solid Blockage,
where velocity around the model increases, because that model (size,
thickness, and thickness distribution) decrease the region during test.
Wake blockage, where velocity around the model increase, because wake
behind model have a signify velocity lesser than the free stream velocity
where the exterior velocity wake in a near wall should be elevated than
the free stream velocity. Although steady amount of stream could
overtake during test, if the exterior wakes have higher velocity so, it has
an inferior pressure and boundary film raise on the model test (Ishak
2006). Total blockage correction can be expressed as:

total = solid + wake (cx)


76

So, uncorrected free stream velocity and dynamic force can be corrected
using by empirical formula:

Vc = Vu (1 + total ) (cxi)

( ( )
qc = qu 1 + 2 M 2 total ) (cxii)

Finally, the aerodynamic characteristic correction can be


articulated by as:

qu
CLc = CLu (cxiii)
qc

qu
CDc = CDu (cxiv)
qc

qu
CMc = CMu (cxv)
qc

We assume incompressible flow with constant temperature, so


correction for temperature, Mach number and pressure neglected.

ii. Buoyancy effect


The buoyancy effect can be expressed as:

dC p S F
C D = (cxvi)
dx SW
77

dC p SF
Where =2,28E -03 (Ishak et al (2006)), =1 if model test without
dx SW
fuselage.

iii. Wall interference correction


Wall interference correction should be done because reflection of the
wing tips vortices in the tunnel walls, floor and ceiling (2006).

iv. STI (Strut, Tare and Interference) corrections


The STI corrections were collected & developed by subtracting all
dummy inverted data from no dummy inverted data that will need to be
subtracted from no dummy upright data for a pitch run only at = -15,
0 and +15 owing to current UTM-LST limitations (Ishak 2006).

In this correction, only effect of blockage is measured as it contributes rather


considerably to the all consequence, the other corrections are unspecified to be very
fewer considerable, and consequently be overlooked (Ishak 2006).

5.2 Free running test

The free running test done because of limited facilities at UTM Towing tank,
because the required maximums speed for testing about 14.4 m/s. The free running
test is divided into two parts, the telemetry data records and calibration RIG.
Telemetry data system will provide results directly, where the data we can get are
rpm versus speed. Calibration test rig was conducted to determine the amount of
thrust for each rpm DC motor. Then the data that is obtained from the telemetry
78

system can be roughly converted usin


usingg the tables generated from the calibration rig.
rig
For more details can be seen from the illustration below (see Figure 5.18 - Figure
5.21). Detail instrumentation set can be seen at Appendix G.

Figure 5.18: Free running test schematic

Figure 5.19: Free running test


79

Figure 5.20: Recording data from running test

Figure 5.21: Real time data dashboard


CHAPTER 6

RESULTS AND DISCUSSION

6.1 Validation Vortex Lattice Method (VLM)

These validations have been done by Adi et.al (2010). The investigation using
a model of NACA 6409 rectangular wing section were performed using Vortex
Lattice Method, where numerical VLM has made using Matlab software (detail
programming can be seen at Appendix) with three different aerodynamic characters
such as aspect Ratios (AR 1, AR 1.5 and AR 2), angle of attack ( = 0o, 2o, 4o, 6o and
8o), Reynolds Number 5 x 106, ground clearance (h/c) 0.1 and 0.3, height endplate
ratio 0.05 c .Ground clearance (h/c ) is ratio of distance between wing trailing edge
and ground surface (h) to wing chord length (c). The number of span wise panel 450
elements and number of chord wise panel 1000, totally number of panel around
450,000 panel elements (see Figure 6.1).
81

Figure 6.1: Panel VLM

6.1.1 Lift coefficient (CL) without endplate

Figure 6.3 - Figure 6.7 shows the comparison results of lift coefficient (CL)
between VLM simulation, CFD simulation (has done by Saeed et al (2010)), and
experimental data (done by Jung et. al. (2008)) for wing only. The validation is
summarized in Table 6.1 Table 6.6. The lift coefficient varied with angle of attack
for three Aspect ratios (AR = 1, 1.5, and 2) and two ground clearances (h/c = 0.1 and
0.3). The influence of aspect ratio on lift coefficient in VLM models would be
declared on CFD simulation and experimental simulation. The magnitude of lift
coefficient increases with increment of aspect ratio, angle of attack, and ground
clearance. According to Figure 6.3 - Figure 6.7 lift coefficients of VLM at 0o - 2o
angle of attack less accurate than CFD results, where CFD closer to experimental
results. At 4o - 8o, VLM results better than CFD results, because near experimental
results.
82

Table 6.1: Lift coefficient versus angle of attack for AR = 1, h/c = 0.1, based on
VLM, CFD, and experimental by Jung et al

Angle of AR = 1
attack VLM CFD Experimental
0 0.357952302 0.254 0.25
2 0.379089462 0.367 0.3
4 0.404755018 0.469 0.356
6 0.460767795 0.561 0.451
8 0.54730893 0.65 0.551

0.7

0.6

0.5

0.4
CL

0.3
VLM
0.2
CFD
0.1
EXPERIMENT JUNG
0
0 2 4 AOA 6 8 10

Figure 6.2: CL versus angle of attack for h/c= 0.1 and AR = 1


83

Table 6.2: Lift coefficient versus angle of attack for AR = 1.5, h/c = 0.1, based on
VLM, CFD, and experimental by Jung et al

Angle of AR = 1.5
attack VLM CFD Experimental
0 0.392493514 0.345 0.27
2 0.460781729 0.491 0.402
4 0.533152467 0.614 0.515
6 0.628637532 0.719 0.603
8 0.745457613 0.82 0.702

0.9
0.8
0.7
0.6
0.5
CL

0.4
VLM
0.3
0.2 CFD
0.1 EXPERIMENT JUNG
0
0 2 4 AOA 6 8 10

Figure 6.3: CL versus angle of attack for h/c= 0.1 and AR = 1.5
84

Table 6.3: Lift coefficient versus angle of attack for AR = 2, h/c = 0.1, based on
VLM, CFD, and experimental by Jung et al

Angle of AR = 2
attack VLM CFD Experimental
0 0.340190829 0.367 0.32
2 0.450304116 0.492 0.45
4 0.564994997 0.612 0.58
6 0.674853409 0.727 0.69
8 0.800041768 0.836 0.8

0.9
0.8
0.7
0.6
0.5
CL

0.4
VLM
0.3
0.2 CFD
0.1 EXPERIMENT JUNG
0
0 2 4 AOA 6 8 10

Figure 6.4: CL versus angle of attack for h/c= 0.1 and AR = 2


85

Table 6.4: Lift coefficient versus angle of attack for AR = 1, h/c = 0.3, based on
VLM, CFD, and experimental by Jung et al

Angle of AR = 1
attack VLM CFD Experimental
0 0.275981291 0.233 0.2
2 0.301317573 0.316 0.28
4 0.347455144 0.395 0.34
6 0.402863156 0.477 0.42
8 0.478528676 0.56 0.51

0.6

0.5

0.4
CL

0.3
VLM
0.2
CFD
0.1
EXPERIMENT JUNG
0
0 2 4 AOA 6 8 10

Figure 6.5: CL versus angle of attack for h/c= 0.3 and AR = 1


86

Table 6.5: Lift coefficient versus angle of attack for AR = 1.5, h/c = 0.3, based on
VLM, CFD, and experimental by Jung et al

Angle of AR = 1.5
attack VLM CFD Experimental
0 0.311971724 0.323 0.27
2 0.366250308 0.432 0.38
4 0.454401137 0.51 0.47
6 0.535782214 0.62 0.54
8 0.635346937 0.723 0.64

0.8
0.7
0.6
0.5
CL

0.4
0.3 VLM
0.2
CFD
0.1
EXPERIMENT JUNG
0
0 2 4 AOA 6 8 10

Figure 6.6: CL versus angle of attack for h/c= 0.3 and AR = 1.5
87

Table 6.6: Lift coefficient versus angle of attack for AR = 2, h/c = 0.3, based on
VLM, CFD, and experimental by Jung et al

Angle of AR = 2
attack VLM CFD Experimental
0 0.42799614 0.417 0.326
2 0.5283454 0.582 0.489
4 0.633181255 0.724 0.624
6 0.756297897 0.842 0.727
8 0.896594578 0.945 0.85

1
0.9
0.8
0.7
0.6
0.5
CL

0.4 VLM
0.3 CFD
0.2
0.1 EXPERIMENT JUNG
0
0 2 4 AOA 6 8 10

Figure 6.7: CL versus angle of attack for h/c= 0.3 and AR = 2


88

6.1.2 Lift coefficient (CL) with endplate

Figure 6.8 Figure 6.13 shows the comparison results of lift coefficient (CL)
between VLM simulation, and experimental data (Jung et. al. (2008)) for wing only
with endplate. The validation is summarized in Table 6.7 - Table 6.12. The lift
coefficient varied with angle of attack for three Aspect ratios (AR = 1, 1.5, and 2),
two ground clearances (h/c = 0.1 and 0.3), and height endplate ratio 0.05c (chord).
The influence of aspect ratio on lift coefficient in VLM models would be declared on
experimental simulation. The magnitude of lift coefficient increases with increment
of aspect ratio, angle of attack, and ground clearance. According to Figure 6.8
Figure 6.13 lift coefficients of VLM at 0o - 4o angle of attack bigger than
experimental results. At 6o - 8o, VLM results less than experimental results.

Table 6.7: Lift coefficient versus angle of attack for AR = 1, h/c = 0.1, endplate
ratio he/c = 0.05, based on VLM, and experimental by Jung et al

Angle of AR=1
attack VLM EXP
0 0.391986373 0.316129032
2 0.415133252 0.367741935
4 0.471442712 0.470967742
6 0.53668419 0.574193548
8 0.637483897 0.709677419
89

Figure 6.8: CL versus AOA for h/c= 0.1 and AR = 1, endplate he/c = 0.05

Table 6.8: Lift coefficient versus angle of attack for AR = 1.5, h/c = 0.1, endplate
ratio he/c = 0.05, based on VLM, and experimental by Jung et al

Angle of AR=1.5
attack VLM EXP
0 0.429811761 0.317460317
2 0.504592812 0.457142857
4 0.583844553 0.615873016
6 0.688408329 0.749206349
8 0.816335653 0.86984127
90

Figure 6.9: CL versus AOA for h/c= 0.1 and AR = 1.5, endplate he/c = 0.1

Table 6.9: Lift coefficient versus angle of attack for AR = 2, h/c = 0.1, endplate
ratio he/c = 0.05, based on VLM, and experimental by Jung et al

Angle of AR=2
attack VLM EXP
0 0.468689973 0.4125
2 0.578580431 0.56875
4 0.693384069 0.69375
6 0.82820663 0.825
8 0.981842706 0.9875
91

Figure 6.10: CL versus AOA for h/c= 0.1 and AR = 2, endplate he/c = 0.05

Table 6.10: Lift coefficient versus angle of attack for AR = 1, h/c = 0.3, endplate
ratio he/c = 0.05, based on VLM, and experimental by Jung et al

Angle of AR=1
attack VLM EXP
0 0.302221566 0.258
2 0.329966819 0.335
4 0.380491146 0.406
6 0.441167347 0.477
8 0.524027137 0.554
92

0.6

0.5

0.4

0.3

CL
0.2
VLM
0.1
EXPERIMENT JUNG
0
0 2 4 AOA 6 8 10

Figure 6.11: CL versus AOA for h/c= 0.3 and AR = 1, endplate he/c = 0.05

Table 6.11: Lift coefficient versus angle of attack for AR = 1.5, h/c = 0.3, endplate
ratio he/c = 0.05, based on VLM, and experimental by Jung et al.

Angle of AR=1.5
attack VLM EXP
0 0.341633967 0.298412698
2 0.401073353 0.4
4 0.497605554 0.501587302
6 0.586724337 0.603174603
8 0.695755664 0.73015873
93

Figure 6.12: CL versus AOA for h/c= 0.3 and AR = 1.5, endplate he/c = 0.05

Table 6.12: Lift coefficient versus angle of attack for AR = 2, h/c = 0.3, endplate
ratio he/c = 0.05, based on VLM, and experimental by Jung et al

Angle of AR=2
attack VLM EXP
0 0.372536141 0.36875
2 0.459882082 0.475
4 0.551133244 0.6
6 0.658296357 0.7
8 0.780413308 0.8375
94

Figure 6.13: CL versus AOA for h/c= 0.3 and AR = 2, endplate he/c = 0.05

6.1.3 Drag coefficient (CD) without endplate

Figure 6.14 - Figure 6.19 shows the comparison results of drag coefficient
(CD) between VLM simulation, CFD simulation (has done by Saeed et al. (2010)),
and experimental data (Jung et. al. (2008)) for wing only. The validation is
summarized in Table 6.13 - Table 6.18. The drag coefficient varied with angle of
attack for three Aspect ratios (AR = 1, 1.5, and 2) and two ground clearances (h/c =
0.1 and 0.3). The influence of aspect ratio on lift coefficient in VLM models would
be declared on CFD simulation and experimental simulation. The magnitude of drag
coefficient increases with increment of aspect ratio, angle of attack, and ground
clearance. According to Figure 6.14 - Figure 6.19 drag coefficients of VLM at almost
all angle of attack more accurate than CFD results, where CFD far away to
experimental results.
95

Table 6.13: Drag coefficient versus angle of attack for AR = 1, h/c = 0.1, based on
VLM, CFD, and experimental by Jung et al

Angle of AR = 1
attack VLM CFD Experimental
0 0.0386313 0.044 0.034
2 0.0394554 0.052 0.037
4 0.049345 0.066 0.043
6 0.0649006 0.084 0.059
8 0.0839587 0.107 0.079

0.12

0.1

0.08
CD

0.06
VLM
0.04
CFD
0.02
EXPERIMENT JUNG
0
0 2 4 AOA 6 8 10

Figure 6.14: CD versus angle of attack for h/c= 0.1 and AR = 1


96

Table 6.14: Drag coefficient versus angle of attack for AR = 1.5, h/c = 0.1, based
on VLM, CFD, and experimental by Jung et al

Angle of AR = 1.5
attack VLM CFD Experimental
0 0.038270726 0.042 0.028
2 0.039918994 0.05 0.031
4 0.050529721 0.065 0.041
6 0.066960894 0.084 0.052
8 0.085503911 0.108 0.071

0.12

0.1

0.08

0.06
CD

0.04 VLM
CFD
0.02
EXPERIMENT JUNG
0
0 2 4 AOA 6 8 10

Figure 6.15: CD versus angle of attack for h/c= 0.1 and AR = 1.5
97

Table 6.15: Drag coefficient versus angle of attack for AR = 2, h/c = 0.1, based on
VLM, CFD, and experimental by Jung et al

Angle of AR = 2
attack VLM CFD Experimental
0 0.038064693 0.039 0.021
2 0.04007352 0.048 0.026
4 0.050941788 0.063 0.034
6 0.067475978 0.082 0.047
8 0.086018994 0.105 0.065

0.12

0.1

0.08

0.06
CD

VLM
0.04
CFD
0.02
EXPERIMENT JUNG
0
0 2 4 AOA 6 8 10

Figure 6.16: CD versus angle of attack for h/c= 0.1 and AR = 2


98

Table 6.16: Drag coefficient versus angle of attack for AR = 1, h/c = 0.3, based on
VLM, CFD, and experimental by Jung et al.

Angle of AR = 1
attack VLM CFD Experimental
0 0.0424303 0.046 0.036
2 0.0433355 0.054 0.039
4 0.0541977 0.065 0.049
6 0.071283 0.081 0.06
8 0.0922153 0.101 0.08

0.12

0.1

0.08
CD

0.06

0.04 VLM

CFD
0.02
EXPERIMENT
JUNG
0
0 2 4 AOA 6 8 10

Figure 6.17: CD versus angle of attack for h/c= 0.3 and AR = 1


99

Table 6.17: Drag coefficient versus angle of attack for AR = 1.5, h/c = 0.3, based
on VLM, CFD, and experimental by Jung et al

Angle of AR = 1.5
attack VLM CFD Experimental
0 0.042034324 0.042 0.0327
2 0.043844685 0.051 0.03758
4 0.055498885 0.062 0.04848
6 0.073545924 0.085 0.06424
8 0.093912487 0.103 0.08121

0.12

0.1

0.08
CD

0.06
VLM
0.04
CFD
0.02
EXPERIMENT JUNG
0
0 2 4 AOA 6 8 10

Figure 6.18: CD versus angle of attack for h/c= 0.3 and AR = 1.5
100

Table 6.18: Drag coefficient versus angle of attack for AR = 2, h/c = 0.3, based on
VLM, CFD, and experimental by Jung et al.

Angle of AR = 2
attack VLM CFD Experimental
0 0.041808029 0.043 0.025
2 0.044014407 0.052 0.03
4 0.055951476 0.066 0.04
6 0.074111662 0.085 0.055
8 0.094478225 0.106 0.075

0.12

0.1

0.08
CD

0.06
VLM
0.04
CFD
0.02
EXPERIMENT JUNG
0
0 2 4 AOA 6 8 10

Figure 6.19: CD versus angle of attack for h/c= 0.3 and AR = 2


101

6.1.4 Drag coefficient (CD) with endplate

Figure 6.20 Figure 6.25

Figure 6.25 shows the comparison results of drag coefficient (CD) between
VLM simulation, and experimental data (Jung et. al. (2008)) for wing only with
endplate. The validation is summarized in

Table 6.19 - Table 6.24. The lift coefficient varied with angle of attack for
three Aspect ratios (AR = 1, 1.5, and 2), two ground clearances (h/c = 0.1 and 0.3),
and height endplate ratio 0.05c (chord). The influence of aspect ratio on lift
coefficient in VLM models would be declared on experimental simulation. The
magnitude of lift coefficient increases with increment of aspect ratio, angle of attack,
and ground clearance. According to Figure 6.20 Figure 6.25 drag coefficients of
VLM at almost all angles are quite similar to the experimental results. Only at AR =
2, h/c = 0.1, and endplate ratio he/c = 0.05 less accurate than experimental results.
102

Table 6.19: Drag coefficient versus angle of attack for AR = 1, h/c = 0.1, endplate
ratio he/c = 0.05, based on VLM, and experimental by Jung et al

Angle of AR=1
attack VLM EXP
0 0.0353822 0.03547
2 0.036137 0.038
4 0.0451949 0.048
6 0.0594421 0.059
8 0.0768973 0.0789

Figure 6.20: CD versus AOA for h/c= 0.1 and AR = 1, endplate he/c = 0.05
103

Table 6.20: Drag coefficient versus angle of attack for AR = 1.5, h/c = 0.1,
endplate ratio he/c = 0.05, based on VLM, and experimental by Jung et al

Angle of AR=1.5
attack VLM EXP
0 0.031913548 0.02788
2 0.033288021 0.03194
4 0.042136191 0.04064
6 0.05583797 0.05297
8 0.071300793 0.07094

Figure 6.21: CD versus AOA for h/c= 0.1 and AR = 1.5, endplate he/c = 0.05
104

Table 6.21: Drag coefficient versus angle of attack for AR = 2, h/c = 0.1, endplate
ratio he/c = 0.05, based on VLM, and experimental by Jung et al

Angle of AR=2
attack VLM EXP
0 0.031741738 0.0235
2 0.033416878 0.0294
4 0.04247981 0.03529
6 0.056267493 0.04706
8 0.071730316 0.06353

Figure 6.22: CD versus AOA for h/c= 0.1 and AR = 2, endplate he/c = 0.05
105

Table 6.22: Drag coefficient versus angle of attack for AR = 1, h/c = 0.3, endplate
ratio he/c = 0.05, based on VLM, and experimental by Jung et al

Angle of AR=1
attack VLM EXP
0 0.0353822 0.03547
2 0.036137 0.038
4 0.0451949 0.048
6 0.0594421 0.059
8 0.0768973 0.0789

0.09

0.08

0.07

0.06

0.05
CD

0.04

0.03

0.02 VLM

0.01 EXPERIMENT JUNG

0
0 2 4 AOA 6 8 10

Figure 6.23: CD versus AOA for h/c= 0.3 and AR = 1, endplate he/c = 0.05
106

Table 6.23: Drag coefficient versus angle of attack for AR = 1.5, h/c = 0.3,
endplate ratio he/c = 0.05, based on VLM, and experimental by Jung et al

Angle of AR=1.5
attack VLM EXP
0 0.035051971 0.03
2 0.036561612 0.036
4 0.046279924 0.045
6 0.061329156 0.059
8 0.078312614 0.0789

Figure 6.24: CD versus AOA for h/c= 0.3 and AR = 1.5, endplate he/c = 0.05
107

Table 6.24: Drag coefficient versus angle of attack for AR = 2, h/c = 0.3, endplate
ratio he/c = 0.05, based on VLM, and experimental by Jung et al

Angle of AR=2
attack VLM EXP
0 0.034863266 0.0341
2 0.036703141 0.03773
4 0.046657335 0.04478
6 0.061800919 0.06125
8 0.078784377 0.078

Figure 6.25: CD versus AOA for h/c= 0.3 and AR = 2, endplate he/c = 0.05
108

6.2 Aerodynamic Trimaran WIG Model

Aerodynamic of Rectangular-Dihedral (ratio 50-50) Wing Trimaran WIG


Model has been calculated using VLM software and has been compared with
experimental result at Wind Tunnel LST-UTM. The investigation was conducted
using NACA 6409 with configuration 50% rectangular wing, 50 % Dihedral reverse
wing with various angle of attack (0o, 2o, 4o, 6o, 8o, 10o) for Aspect Ratios (AR) 1.25,
various ground clearances (from h/c = 0 until h/c = 0.4), and taper 0.8 with dihedral
angle 13o at dihedral reverse wing without endplate and without endplate.

6.2.1 Numerical Results

Figure 6.26 show results of lift coefficient (CL) from VLM simulation, for
Rectangular-Dihedral (50-50) Wing Trimaran WIG Model without endplate. The
result is summarized in Table 6.25. The lift coefficient varied with angle of attack for
Aspect Ratios (AR = 1.25), and varied with ground clearances (from h/c = 0 until h/c
= 0.4). The magnitude of lift coefficient decreases with increment of ground
clearance, otherwise lift coefficient increase with increment angle of attack. Figure
6.27 show results of lift coefficient (CL) from VLM simulation, for Rectangular-
Dihedral (50-50) Wing Trimaran WIG Model with endplate 0.05c. The result is
summarized in Table 6.26 with various angle of attack for Aspect Ratios (AR =
1.25), and various ground clearances (from h/c = 0 until h/c = 0.4). As well as at
Figure 6.27, the magnitude of lift coefficient decreases with increment of ground
clearance, and lift coefficient increase with increment angle of attack. If the result
Figure 6.26 and Figure 6.27 is compared, lift coefficients wing with endplate
increase around 14.5%.
109

Figure 6.26: CL versus h/c every AOA for AR = 1.25 rectangular-dihedral


(50-50) wing trimaran WIG model without endplate using VLM

Table 6.25: Lift coefficient every angle of attack (AOA) versus ground clearance
(h/c) rectangular-dihedral (50-50) wing trimaran WIG model without endplate (AR =
1.25) using VLM

0 DEGREE AOA 2 DEGREE AOA 4 DEGREE AOA 6 DEGREE AOA 8 DEGREE AOA 10 DEGREE AOA

h/c CL h/c CL h/c CL h/c CL h/c CL h/c CL


0.02 0.620801 0.02 0.699522 0.02 0.784827 0.02 0.913803 0.02 1.084022 0.02 1.28236
0.04 0.493606 0.04 0.556198 0.04 0.624025 0.04 0.726575 0.04 0.861918 0.04 1.019619
0.06 0.433768 0.06 0.488772 0.06 0.548377 0.06 0.638495 0.06 0.757431 0.06 0.896015
0.08 0.398187 0.08 0.448679 0.08 0.503395 0.08 0.586121 0.08 0.695301 0.08 0.822517
0.1 0.374366 0.1 0.421838 0.1 0.47328 0.1 0.551057 0.1 0.653705 0.1 0.773311
0.12 0.357213 0.12 0.40251 0.12 0.451595 0.12 0.525809 0.12 0.623754 0.12 0.737879
0.14 0.344234 0.14 0.387885 0.14 0.435187 0.14 0.506704 0.14 0.60109 0.14 0.711069
0.16 0.334051 0.16 0.376411 0.16 0.422313 0.16 0.491715 0.16 0.583309 0.16 0.690034
0.18 0.325838 0.18 0.367156 0.18 0.411931 0.18 0.479626 0.18 0.568968 0.18 0.67307
0.2 0.319069 0.2 0.359528 0.2 0.403372 0.2 0.469661 0.2 0.557147 0.2 0.659086
0.22 0.313389 0.22 0.353128 0.22 0.396192 0.22 0.461301 0.22 0.54723 0.22 0.647354
0.24 0.308554 0.24 0.34768 0.24 0.390079 0.24 0.454183 0.24 0.538786 0.24 0.637365
0.26 0.304386 0.26 0.342983 0.26 0.38481 0.26 0.448048 0.26 0.531508 0.26 0.628755
0.28 0.300755 0.28 0.338892 0.28 0.38022 0.28 0.442703 0.28 0.525168 0.28 0.621256
0.3 0.297563 0.3 0.335296 0.3 0.376185 0.3 0.438005 0.3 0.519595 0.3 0.614663
0.32 0.294735 0.32 0.332109 0.32 0.372609 0.32 0.433842 0.32 0.514656 0.32 0.60882
0.34 0.292211 0.34 0.329265 0.34 0.369418 0.34 0.430127 0.34 0.510249 0.34 0.603607
0.36 0.289945 0.36 0.326711 0.36 0.366553 0.36 0.426791 0.36 0.506292 0.36 0.598926
0.38 0.287898 0.38 0.324405 0.38 0.363966 0.38 0.423779 0.38 0.502719 0.38 0.594699
0.4 0.286041 0.4 0.322313 0.4 0.361618 0.4 0.421045 0.4 0.499476 0.4 0.590862
110

Figure 6.27: CL versus h/c every AOA for AR = 1.25 rectangular-dihedral (50-50)
wing trimaran WIG model with endplate 0.05c using VLM

Table 6.26: Lift coefficient every angle of attack (AOA) versus ground clearance
(h/c) rectangular-dihedral (50-50) wing trimaran WIG model (AR = 1.25) with
endplate (he/c=0.05) using VLM

0 DEGREE AOA 2 DEGREE AOA 4 DEGREE AOA 6 DEGREE AOA 8 DEGREE AOA 10 DEGREE AOA

h/c CL h/c CL h/c CL h/c CL h/c CL h/c CL


0.02 0.7164806 0.02 0.807334 0.02 0.9057874 0.02 1.0546409 0.02 1.2510944 0.02 1.2510944
0.04 0.5696819 0.04 0.6419205 0.04 0.7202019 0.04 0.838557 0.04 0.9947594 0.04 0.9947594
0.06 0.5006217 0.06 0.5641031 0.06 0.6328948 0.06 0.7369022 0.06 0.8741688 0.06 0.8741688
0.08 0.459557 0.08 0.5178312 0.08 0.5809801 0.08 0.676456 0.08 0.802463 0.08 0.802463
0.1 0.4320644 0.1 0.4868524 0.1 0.5462235 0.1 0.6359876 0.1 0.7544564 0.1 0.7544564
0.12 0.4122682 0.12 0.4645459 0.12 0.5211967 0.12 0.6068481 0.12 0.7198889 0.12 0.7198889
0.14 0.3972885 0.14 0.4476667 0.14 0.5022591 0.14 0.5847984 0.14 0.6937319 0.14 0.6937319
0.16 0.3855361 0.16 0.4344241 0.16 0.4874016 0.16 0.5674992 0.16 0.6732102 0.16 0.6732102
0.18 0.3760577 0.18 0.4237437 0.18 0.4754187 0.18 0.5535471 0.18 0.6566593 0.18 0.6566593
0.2 0.3682447 0.2 0.41494 0.2 0.4655414 0.2 0.5420466 0.2 0.6430165 0.2 0.6430165
0.22 0.3616896 0.22 0.4075537 0.22 0.4572544 0.22 0.5323978 0.22 0.6315703 0.22 0.6315703
0.24 0.3561088 0.24 0.4012652 0.24 0.450199 0.24 0.5241829 0.24 0.6218252 0.24 0.6218252
0.26 0.3512984 0.26 0.3958448 0.26 0.4441176 0.26 0.5171021 0.26 0.6134255 0.26 0.6134255
0.28 0.3471081 0.28 0.3911232 0.28 0.4388202 0.28 0.5109341 0.28 0.6061085 0.28 0.6061085
0.3 0.3434245 0.3 0.3869725 0.3 0.4341633 0.3 0.5055119 0.3 0.5996763 0.3 0.5996763
0.32 0.3401604 0.32 0.3832945 0.32 0.4300368 0.32 0.5007073 0.32 0.5939766 0.32 0.5939766
0.34 0.3372476 0.34 0.3800123 0.34 0.4263544 0.34 0.4964197 0.34 0.5888904 0.34 0.5888904
0.36 0.334632 0.36 0.3770651 0.36 0.4230477 0.36 0.4925697 0.36 0.5843232 0.36 0.5843232
0.38 0.3322702 0.38 0.3744038 0.38 0.4200619 0.38 0.4890932 0.38 0.5801991 0.38 0.5801991
0.4 0.3301268 0.4 0.3719885 0.4 0.4173521 0.4 0.485938 0.4 0.5764563 0.4 0.5764563
111

Figure 6.28 show results of drag coefficient (CD) from VLM simulation, for
Rectangular-Dihedral (50-50) Wing Trimaran WIG Model without endplate. The
result is summarized in Table 6.27. The drag coefficient varied with angle of attack
for Aspect Ratios (AR = 1.25), and varied with ground clearances (from h/c = 0 until
h/c = 0.4). The magnitude of drag coefficient increases with increment of ground
clearance, and drag coefficient increase with increment angle of attack. Figure 6.29
show results of drag coefficient (CD) from VLM simulation, for Rectangular-
Dihedral (50-50) Wing Trimaran WIG Model with endplate 0.05c. The result is
summarized in Table 6.28 with various angle of attack for Aspect Ratios (AR =
1.25), and various ground clearances (from h/c = 0 until h/c = 0.4). The magnitude of
drag coefficient increases with increment of ground clearance, and drag coefficient
increase with increment angle of attack. Like above comparison, if the result Figure
6.28 and Figure 6.29 is compared, drag coefficients wing with endplate decrease
around 25.17%.

Figure 6.28: CD versus h/c every AOA for AR = 1.25 rectangular-dihedral


(50-50) wing trimaran WIG model without endplate using VLM
112

Table 6.27: Drag coefficient every angle of attack (AOA) versus ground clearance
(h/c) rectangular-dihedral (50-50) wing trimaran WIG model (AR = 1.25) without
endplate using VLM

0 DEGREE AOA 2 DEGREE AOA 4 DEGREE AOA 6 DEGREE AOA 8 DEGREE AOA 10 DEGREE AOA

h/c CD h/c CD h/c CD h/c CD h/c CD h/c CD


0.02 0.0139735 0.02 0.0144605 0.02 0.0182067 0.02 0.023976 0.02 0.0309065 0.02 0.0387736
0.04 0.0221029 0.04 0.0228732 0.04 0.0287989 0.04 0.0379245 0.04 0.0488871 0.04 0.0613311
0.06 0.0286216 0.06 0.0296192 0.06 0.0372925 0.06 0.0491095 0.06 0.0633052 0.06 0.0794193
0.08 0.0339653 0.08 0.035149 0.08 0.044255 0.08 0.0582782 0.08 0.0751242 0.08 0.0942468
0.1 0.0384253 0.1 0.0397645 0.1 0.0500661 0.1 0.0659307 0.1 0.0849888 0.1 0.1066223
0.12 0.042204 0.12 0.043675 0.12 0.0549897 0.12 0.0724144 0.12 0.0933467 0.12 0.1171077
0.14 0.0454466 0.14 0.0470306 0.14 0.0592147 0.14 0.0779781 0.14 0.1005187 0.14 0.1261053
0.16 0.0482596 0.16 0.0499416 0.16 0.0628798 0.16 0.0828047 0.16 0.1067404 0.16 0.1339107
0.18 0.050723 0.18 0.0524908 0.18 0.0660895 0.18 0.0870314 0.18 0.1121889 0.18 0.1407461
0.2 0.0528982 0.2 0.0547418 0.2 0.0689236 0.2 0.0907636 0.2 0.117 0.2 0.1467818
0.22 0.0548329 0.22 0.056744 0.22 0.0714445 0.22 0.0940833 0.22 0.1212793 0.22 0.1521504
0.24 0.0565651 0.24 0.0585365 0.24 0.0737014 0.24 0.0970553 0.24 0.1251104 0.24 0.1569567
0.26 0.0581248 0.26 0.0601506 0.26 0.0757336 0.26 0.0997315 0.26 0.1285602 0.26 0.1612846
0.28 0.0595366 0.28 0.0616116 0.28 0.0775732 0.28 0.102154 0.28 0.1316829 0.28 0.1652022
0.3 0.0608207 0.3 0.0629404 0.3 0.0792462 0.3 0.1043572 0.3 0.1345229 0.3 0.1687651
0.32 0.0619935 0.32 0.0641541 0.32 0.0807744 0.32 0.1063696 0.32 0.137117 0.32 0.1720195
0.34 0.063069 0.34 0.0652671 0.34 0.0821757 0.34 0.1082149 0.34 0.1394958 0.34 0.1750038
0.36 0.0640588 0.36 0.0662914 0.36 0.0834653 0.36 0.1099132 0.36 0.141685 0.36 0.1777502
0.38 0.0649727 0.38 0.0672372 0.38 0.0846561 0.38 0.1114813 0.38 0.1437064 0.38 0.1802862
0.4 0.0658192 0.4 0.0681131 0.4 0.085759 0.4 0.1129337 0.4 0.1455786 0.4 0.1826349

Figure 6.29: CD versus h/c every AOA for AR = 1.25 rectangular-dihedral


(50-50) wing trimaran WIG model with endplate 0.05c using VLM
113

Table 6.28: Drag coefficient every angle of attack (AOA) versus ground clearance
(h/c) rectangular-dihedral (50-50) wing trimaran WIG model (AR = 1.25) with
endplate (he/c=0.05), using VLM

0 DEGREE AOA 2 DEGREE AOA 4 DEGREE AOA 6 DEGREE AOA 8 DEGREE AOA 10 DEGREE AOA

h/c CL h/c CL h/c CL h/c CL h/c CL h/c CL


0.02 0.0104906 0.02 0.0108562 0.02 0.0136687 0.02 0.018 0.02 0.0232031 0.02 0.0232031
0.04 0.0165938 0.04 0.0171721 0.04 0.0216208 0.04 0.0284719 0.04 0.036702 0.04 0.036702
0.06 0.0214877 0.06 0.0222366 0.06 0.0279974 0.06 0.036869 0.06 0.0475264 0.06 0.0475264
0.08 0.0254994 0.08 0.0263882 0.08 0.0332245 0.08 0.0437524 0.08 0.0563996 0.08 0.0563996
0.1 0.0288478 0.1 0.0298532 0.1 0.0375872 0.1 0.0494975 0.1 0.0638054 0.1 0.0638054
0.12 0.0316847 0.12 0.032789 0.12 0.0412836 0.12 0.0543652 0.12 0.0700801 0.12 0.0700801
0.14 0.0341191 0.14 0.0353082 0.14 0.0444554 0.14 0.0585421 0.14 0.0754645 0.14 0.0754645
0.16 0.0362309 0.16 0.0374937 0.16 0.047207 0.16 0.0621657 0.16 0.0801354 0.16 0.0801354
0.18 0.0380803 0.18 0.0394075 0.18 0.0496167 0.18 0.0653389 0.18 0.0842259 0.18 0.0842259
0.2 0.0397133 0.2 0.0410975 0.2 0.0517445 0.2 0.0681409 0.2 0.0878378 0.2 0.0878378
0.22 0.0411659 0.22 0.0426006 0.22 0.053637 0.22 0.0706331 0.22 0.0910505 0.22 0.0910505
0.24 0.0424663 0.24 0.0439463 0.24 0.0553314 0.24 0.0728644 0.24 0.0939267 0.24 0.0939267
0.26 0.0436372 0.26 0.0451581 0.26 0.0568571 0.26 0.0748735 0.26 0.0965167 0.26 0.0965167
0.28 0.0446972 0.28 0.046255 0.28 0.0582381 0.28 0.0766922 0.28 0.098861 0.28 0.098861
0.3 0.0456612 0.3 0.0472526 0.3 0.0594942 0.3 0.0783462 0.3 0.1009932 0.3 0.1009932
0.32 0.0465417 0.32 0.0481638 0.32 0.0606414 0.32 0.079857 0.32 0.1029407 0.32 0.1029407
0.34 0.0473491 0.34 0.0489993 0.34 0.0616935 0.34 0.0812424 0.34 0.1047266 0.34 0.1047266
0.36 0.0480922 0.36 0.0497683 0.36 0.0626617 0.36 0.0825174 0.36 0.1063701 0.36 0.1063701
0.38 0.0487783 0.38 0.0504783 0.38 0.0635556 0.38 0.0836947 0.38 0.1078877 0.38 0.1078877
0.4 0.0494138 0.4 0.051136 0.4 0.0643836 0.4 0.084785 0.4 0.1092932 0.4 0.1092932

6.2.2 Experimental Results

Figure 6.30 show results of lift coefficient (CL) from wind tunnel test LST-
UTM, for Rectangular-Dihedral (50-50) Wing Trimaran WIG Model with endplate.
The result is summarized in Table 6.29. The lift coefficient varied with angle of
attack for Aspect Ratios (AR = 1.25), and with varied ground clearances (h/c = 0.06,
0.1, 0.2, 0.4). The magnitude of lift coefficient decreases with increment of ground
clearance, otherwise lift coefficient increase with increment angle of attack.
114

Table 6.29: Experimental result of lift coefficient every angle of attack (AOA)
versus ground clearance (h/c) rectangular-dihedral (50-50) wing trimaran WIG
model with endplate (AR = 1.25)

h/c 0 deg 2 deg 4 deg 6 deg 8 deg 10 deg


0.06 0.274 0.382 0.509 0.657 0.814 0.992
0.1 0.251 0.360 0.486 0.633 0.791 0.966
0.2 0.193 0.306 0.431 0.573 0.733 0.901
0.4 0.076 0.200 0.319 0.452 0.618 0.772

Figure 6.30: Experimental result of lift coefficient every angle of attack (AOA)
versus ground clearance (h/c) rectangular-dihedral (50-50) wing trimaran WIG
Model with endplate (AR = 1.25)

Figure 6.31 show results of drag coefficient (CD) from wind tunnel test LST-
UTM, for Rectangular-Dihedral (50-50) Wing Trimaran WIG Model without
endplate. The result is summarized in Table 6.30. The drag coefficient varied with
115

angle of attack for Aspect Ratios (AR = 1.25), and varied with ground clearances
(h/c = 0.06, 0.1, 0.2, 0.4). The magnitude of drag coefficient increases with
increment of ground clearance, and drag coefficient increase with increment angle of
attack.

Table 6.30: Experimental result of drag coefficient every angle of attack (AOA)
versus ground clearance (h/c) rectangular-dihedral (50-50) wing trimaran WIG
model with endplate (AR = 1.25)
h/c 0 deg 2 deg 4 deg 6 deg 8 deg 10 deg
0.06 0.025009 0.026429 0.0394133 0.051727 0.063488 0.07634
0.1 0.028218 0.0305842 0.0427602 0.054590 0.06711 0.08061
0.2 0.036239 0.0409713 0.051127 0.058167 0.071648 0.09129
0.4 0.052281 0.0617455 0.067862 0.0760567 0.09429 0.11264

Figure 6.31: Experimental result of drag coefficient every angle of attack (AOA)
versus ground clearance (h/c) rectangular-dihedral (50-50) wing trimaran WIG
model with endplate (AR = 1.25)
116

Based on the chapter fifth "experimental work", there are several corrections
that must be done to improve the result generated from the experiments. A very
significant correction that must be done is Blockage effect. In this experiment, only
wing (without fuselage) that has been tested in wind tunnel test, so buoyancy effect
correction negligible. Ground effect resulting from wall is needed in the calculations,
so wall interference effect correction is also ignored. .STI correction is ignored
because it does not significantly affect the results. Blockage effect has corrected by
equation (cxiii) and equation (cxiv).

6.2.3 Comparison numerical results with experimental result

Figure 6.32 Figure 6.34 shows the comparison results of lift coefficient (CL)
between VLM simulation, Experimental LST-UTM, and Experimental LST-UTM
after correction for wing only with endplate. The validation is summarized in

Table 6.31 - Table 6.33. The lift coefficient varied with angle of attack for
Aspect ratios (AR = 1.25), three ground clearances (h/c = 0.06, 0.1 and 0.15), and
height endplate ratio 0.06c (chord). The magnitude of lift coefficient increases with
increment of aspect ratio, angle of attack, and ground clearance. According all table,
lift coefficients of VLM at 0o - 8o angle of attack bigger than experimental results.
The differences that occur are 27.7% - 37.4% at 0o, 15.11% 15.9% at 2o, 7.55% -
8.9 % at 4o, 2.3% - 4.55% at 6o, 0.04% - 0.8% at 8o. While the blockage effect
correction only reduce experiment result approximately 0.01% - 4.7%.
117

Table 6.31: Lift coefficient versus angle of attack for AR = 1.25, h/c = 0.06,
endplate ratio he/c = 0.06, based on VLM, experimental LST-UTM, and
experimental LST-UTM after correction.

Angle of AR = 1.25
attack VLM Experiment Exp. After Correction
0 0.376057666 0.274335 0.271591697
2 0.447666724 0.38213 0.380008174
4 0.546223472 0.508853 0.503224784
6 0.676455994 0.656465 0.645618792
8 0.802463006 0.813657 0.795767884

Figure 6.32: Comparison of CL versus AOA for h/c= 0.06 he/c=0.06 and AR =
1.25 between VLM, experimental LST-UTM, and experimental LST-UTM after
correction
118

Table 6.32: Lift coefficient versus angle of attack for AR = 1.25, h/c = 0.1,
endplate ratio he/c = 0.06, based on VLM, experimental LST-UTM, and
experimental LST-UTM after correction

AR = 1.25
Angle of attack Exp. After
VLM Experiment Correction
0 0.361689632 0.2510218 0.248511544
2 0.423743704 0.3606635 0.356188301
4 0.521196723 0.4864629 0.474521992
6 0.635987638 0.6324594 0.60942323
8 0.754456397 0.7905743 0.753711715

Figure 6.33: Comparison of CL versus AOA for h/c= 0.1 he/c=0.06 and AR = 1.25
between VLM, experimental LST-UTM, and experimental LST-UTM after
correction
119

Table 6.33: Lift coefficient versus angle of attack for AR = 1.25, h/c = 0.15,
endplate ratio he/c = 0.06, based on VLM, experimental LST-UTM, and
experimental LST-UTM after correction

AR = 1.25
Angle of attack Exp. After
VLM Experimental Correction
0 0.351298402 0.2218802 0.219661354
2 0.391123177 0.3338301 0.330942616
4 0.487401556 0.4584754 0.450601124
6 0.606848094 0.6024522 0.592498258
8 0.719888876 0.75018 0.719596239

Figure 6.34: Comparison of CL versus AOA for h/c= 0.15 he/c=0.06 and AR =
1.25 between VLM, experimental LST-UTM, and experimental LST-UTM after
correction

Figure 6.35 - Figure 6.37 shows the comparison results of drag coefficient
(CL) between VLM simulation, Experimental LST-UTM, and Experimental LST-
UTM after correction for wing only with endplate. The validation is summarized in
120

Table 6.34 - Table 6.36. The lift coefficient varied with angle of attack for Aspect
ratios (AR = 1.25), three ground clearances (h/c = 0.06, 0.1 and 0.15), and height
endplate ratio 0.06c (chord). The magnitude of drag coefficient increases with
increment of aspect ratio, angle of attack, and ground clearance. According to Figure
6.35 - Figure 6.37 lift coefficients of VLM approximately at 0o angle of attack bigger
than experimental results. But at 2o - 8o angle of attack less than experimental results,
difference case occur at h/c= 0.15 almost drag coefficient VLM result like
experimental results. The differences that occur are 2.9% - 6.45% at 0o, -0.3% -
1.5% at 2o, -2.7% -17.3 % at 4o, -16.3% 2.8% at 6o, -4.08% - 8.9% at 8o. While
the blockage effect correction only reduce experiment result approximately 0.01% -
4.1%.

Table 6.34: Drag coefficient versus angle of attack for AR = 1.25, h/c = 0.06,
endplate ratio he/c = 0.06, based on VLM, experimental LST-UTM, and
experimental LST-UTM after correction

Angle of AR = 1.25
attack VLM Experimental Exp. After Correction
0 0.0254994 0.0250098 0.024759748
2 0.0263882 0.0264294 0.026282655
4 0.0332245 0.0394133 0.038977379
6 0.0437524 0.0517278 0.050873126
8 0.0563996 0.0600259 0.058706224
121

Figure 6.35: Comparison of CD versus AOA for h/c= 0.15 he/c=0.06 and AR =
1.25 between VLM, experimental LST-UTM, and experimental LST-UTM after
correction

Table 6.35: Drag coefficient versus angle of attack for AR = 1.25, h/c = 0.1,
endplate ratio he/c = 0.06, based on VLM, experimental LST-UTM, and
experimental LST-UTM after correction

Angle of attack AR = 1.25


Exp. After
VLM Experimental Correction
0 0.028847786 0.0282183 0.027936125
2 0.029853205 0.0305843 0.030319721
4 0.037587196 0.0392979 0.038622954
6 0.049497542 0.05459 0.053688061
8 0.063805426 0.0671123 0.064376187
122

Figure 6.36: Comparison of CD versus AOA for h/c= 0.15 he/c=0.06 and AR =
1.25 between VLM, experimental LST-UTM, and experimental LST-UTM after
correction

Table 6.36: Drag coefficient versus angle of attack for AR = 1.25, h/c = 0.15,
endplate ratio he/c = 0.06, based on VLM, experimental LST-UTM, and
experimental LST-UTM after correction

AR = 1.25
Angle of attack Exp. After
VLM Experimental Correction
0 0.034119095 0.0322289 0.031906597
2 0.035308232 0.0357778 0.035468353
4 0.044455443 0.046944 0.046137699
6 0.058542147 0.0581678 0.057206728
8 0.075464487 0.0716422 0.068721441
123

Figure 6.37: Comparison of CD versus AOA for h/c= 0.15 he/c=0.06 and AR =
1.25 between VLM, experimental LST-UTM, and experimental LST-UTM after
correction

6.3 Hydrodynamic Trimaran WIG Model

Hydrodynamic trimaran WIG model was estimated using Savitsky method


program, where this program was developed using Savitsky formula and other
numerical and empirical formula. This program was validated with commercial
software Maxsurf Hullspeed.
124

6.3.1 Result and validation Savitsky Method (SM) program with Hullspeed
Maxsurf.

These numerical simulations were performed using Savitsky Method (SM),


where numerical SM was modified with clement formula (0, 0, 0) to calculate
planning hull with step. The simulation was conducted with step 5% in front of
midship, speed around 0- 28 knot. These numerical simulations was performed
comparison between hull without steps and hull with step (clement steps) (Figure
6.38) , clear clement step hull produce lesser wake compare to the hull without step
which indicates clement hulls efficiency in terms of reducing hull resistance (Figure
6.39). These simulations have been validated with commercial Software Hullspeed
Maxsurf by noverdo et.al (2010) for calculate Trimaran WIG Prototype (Figure
6.40). From Figure 6.40 show the comparison results of drag water resistance
between SM simulation and Hullspeed Maxsurf for planning condition only, its
difference about 8.7 %. Detail simulation can be seen at Appendix F.

Figure 6.38: Hull without step and hull with step (Clementss step) (2010)
125

Figure 6.39: Planing hull simulation using Hullspeed (2010)

2000

1800

1600

1400

1200
DRAG WATER (N)

1000

800

600

400

200
SM
HULLSPEED
0
0 10 20 30 40 50
SPEED(KNOT)

Figure 6.40: SM & Hullspeed Maxsurf comparison [0]


126

Finally, hydrodynamic Trimaran WIG Model without aerodynamic effect was


investigated without or with step in Figure 6.41 and was summarized at Table 6.37.
Figure 6.41 shows thrust WIG trimaran model without step increase from 0-4 knot,
but when entering in speed 5 knot thrust decrease. This is hump phenomena, where
WIG trimaran model phase change from not-planning condition into planning
condition, after that thrust increase. Different with thrust WIG trimaran model with
Clementss step, after entering phase planning hull, thrust decrease, cause Clementss
step reduce spray resistance planing hull, for detail illustration see Figure 6.39. In
this simulation also show the comparison results of lift coefficient (CL) between flat
planing surface and vee planing surface (see Figure 6.42), where for trimaran WIG
model using vee planing surface.

Table 6.37: Drag water resistance trimaran WIG model with and without Step

V(knot) Step No Step


0 0 0
2 0.205746 0.226796
3 0.411492 0.453592
4 0.696039 0.994002
5 0.739101 0.931861
6 0.795435 0.912511
8 0.921248 0.998557
10 0.957213 1.137501
12 0.883638 1.283672
14 0.773276 1.458618
16 0.666103 1.677205
18 0.573435 1.944325
20 0.496157 2.261427
22 0.432272 2.629338
25 0.356311 3.279654
26 0.335243 3.52356
28 0.298167 4.05462
127

Figure 6.41: Drag water resistance trimaran WIG model with or without step using
Savitsky Method

Figure 6.42: Lift coefficient planing surface with and without step
128

6.3.2 Free running test result

According with the methodology of research and experimental work


procedure, we must to do calibration rigs in advance to know the thrust generated
each round of his DC motor. Figure 6.43 show result calibration rig, where
experiment was starting from 2085 rpm until 9900 rpm and result come out from
0.18 kg until 2.6 kg. The result is summarized in Table 6.38.

Table 6.38: Calibration RIG result

RPM THRUST(kg)
2085 0.18
2385 0.295
2475 0.395
3060 0.6
3175 0.69
3895 0.7
4170 0.883109
4770 1.067729
4950 1.123115
6120 1.483124
6350 1.553895
7790 1.996983
8340 2.166218
9540 2.535458
9900 2.64623
129

2.5

Thrust (kg)
1.5

0.5 THRUST(kg)
Linear
0
0 5000 10000 15000
RPM

Figure 6.43: Calibration RIG result (Thrust versus RPM)

After doing the calibration rig and then we started to retrieve data using data
telemetry system. Figure 6.44 shows the data obtained directly when the experiments
conducted, where data is obtained rpm vs. speed, and then using the Table 6.38, we
can determine how much thrust was generated every each propulsion in use. Free
running test uses 2 EDF (Electric Ducting Propeller) so that the thrust can be
multiplied 2 according with amount of propulsion in experiment. For calculate total
drag we using formula from Fox & Blound (1976). Figure 6.45 represents the results
obtained from the telemetry system after process. All result and process was
summarized at Table 6.39.
130

Figure 6.44: Free running dashboard

Table 6.39
39: Thrust from experiment free running test

Thrust 1
Thrust 2 Total
V(knot) RPM motor
motor(kg) Drag(T/Mfactor)
(kg)
0 0 0 0 0
2 2025 0.2735 0.547 0.558163265
3 2379 0.3797 0.7594 0.774897959
4 2445 0.3995 0.799 0.815306122
5 3045 0.5795 1.159 1.182653061
6 3165 0.6155 1.231 1.256122449
8 4030 0.875 1.75 1.785714286
10 3895 0.8345 1.669 1.703061224
12 4768 1.0964 2.1928 2.23755102
14 4788 1.1024 2.2048 2.249795918
16 5667 1.3661 2.7322 2.787959184
18 5789 1.4027 2.8054 2.862653061
20 5987 1.4621 2.9242 2.983877551
22 6103 1.4969 2.9938 3.054897959
25 6456 1.6028 3.2056 3.271020408
26 6678 1.6694 3.3388 3.406938776
28 6979 1.7597 3.5194 3.59122449
131

4
3.5
3

Total Drag (kg)


2.5
2
1.5
1
0.5
0
0 5 10 15 20 25 30
RPM

Figure 6.45: Total drag trimaran WIG model result from free running test

6.3.3 Comparison numerical result (SM) with free running test

Figure 6.46 are shows comparison Thrust Trimaran WIG model between SM
with step, SM without step and free running test on speed range from 0 until 28 knot.
According Figure 6.46 trend line from free running test really different with trend
line from SM with step, but looks similar with without step. Its mean no more effect
stepped hull in this model, because step in this model very small. So in this moment
we assume Trimaran WIG model without stepped hull. At 0 5 knot numerical
result (SM without step) near with experimental (free running test) result (not
planing hull phase), where aerodynamic drag and side hull not more effect in this
result, but from 6-24 knot experimental result bigger than numerical result (SM
without step). This phenomena was happens because of aerodynamic drag (air drag
wing and fuselage) began to affect of experiment results, because numerical result
(SM without step) only for the hydrodynamic drag, and therefore need to be added
aerodynamic drag on the numerical results (scope other master student). From 24
28 knot numerical results (SM without step) bigger than experimental results (free
132

running test). This moment was occurred because draft fuselage Trimaran WIG was
changed or all body Trimaran WIG was began fly (take off condition). For detail
condition we can see Figure 6.47. The comparison was summarized in Table 6.40.

4.5
4
SM with step
3.5 hull
3
2.5
2
1.5
1
0.5
0
0 5 10 15 20 25 30

Figure 6.46: Comparison Drag Trimaran Wig Model every speed between
numerical result (SM) and experimental work (free running test)

Gap 1 cm (normal condition) Gap 2.5 cm

Gap 4 cm Gap 5 cm

Figure 6.47: Draft change during take off


133

Table 6.40: Comparison drag trimaran WIG model between SM and free running test

SM
SM with
V without EXP
step
step
0 0 0 0
2 0.205746 0.226796 0.558163
3 0.411492 0.453592 0.774898
4 0.696039 0.994002 0.815306
5 0.739101 0.931861 1.182653
6 0.795435 0.912511 1.256122
8 0.921248 0.998557 1.785714
10 0.957213 1.137501 1.703061
12 0.883638 1.283672 2.237551
14 0.773276 1.458618 2.249796
16 0.666103 1.677205 2.787959
18 0.573435 1.944325 2.862653
20 0.496157 2.261427 2.983878
22 0.432272 2.629338 3.054898
25 0.356311 3.279654 3.27102
26 0.335243 3.52356 3.406939
28 0.298167 4.05462 3.591224

6.4 Longitudinal Static Stability Trimaran WIG Model

These simulations were conducted to investigate the Static Stability Margin


(SSM) during take-off rectangular-dihedral reverse wing NACA 6409 Trimaran WIG
model, aspect ratio (AR) 1.25, ground clearance (h/c) = 0.06, 0.1, 0.3, endplate he/c
= 0.06c. Regarding equation (lxxiv) the distance between the Aerodynamic Centre in
Pitch (ACP), Hydrodynamic Centre in Heave (HCH) and Aerodynamic Centre in
Height (ACH) must be bigger than zero. Figure 6.48 show stability condition wing
without endplate in variation ground clearance. At h/c = 0.06, SSM in stable
condition, but when entering at h/c 0.1, SSM becomes unstable, and after entering
h/c = 0.3, SSM becomes stable again. The phenomenon occurred again in WIG with
endplate (Figure 6.49). From both of figures it can be concluded there is not a stable
phase between h/c = 0.06 and h/c 0.3, where these phase is a take-off phase. The
results are summarized in Table 6.41 Table 6.42. To overcome instability during
134

the takeoff, tail added (NACA 0012, AR = 5) at the trimaran design WIG. Figure
6.50 show stability condition wing with endplate in variation ground clearance after
added tail. SSM become stable in all ground clearance (h/c). The results are
summarized in Table 6.43.

Table 6.41: SSM on rectangular-dihedral reverse wing NACA 6409 A/R = 1.25
trimaran WIG model without endplate

WITHOUT ENDPLATE
ANALYSIS STATIC STABILITY MARGIN
(assume cog=0.3 c)
h/c h/c h/c
0.06 0.1 0.3
Cma -0.012 -0.0103 -0.0082
Cla 0.0399 0.0344 0.0273
Cmz 1.3412 0.8437 0.2409
Clz -4.4708 -2.8125 -0.803
CmH 0.0000001 0 0
ClH -0.001 0 0
Xa 0.60075188 0.5994186 0.6003663
Xz 0.59999105 0.59998222 0.6
XH -0.0001 0 0
SSM 0.00066083 -0.0005636 0.0003663
stable not stable stable

Figure 6.48: SSM wing without endplate, h/c= 0.06 - 0.3


135

Table 6.42: SSM on rectangular-dihedral reverse wing NACA 6409 A/R = 1.25
trimaran WIG model with endplate (0.06c)

WITH ENDPLATE(0.06c)
ANALYSIS STATIC STABILITY MARGIN
(assume cog=0.3 c)

h/c h/c h/c


0.06 0.1 0.3
Cma -0.0138 -0.0119 -0.0095
Cla 0.046 0.0506 0.0316
Cmz 1.548 0.9738 0.274
Clz -5.1599 -3.3948 -0.9692
CmH 0.00000013 0 0
ClH -0.001 0 0
Xa 0.601 0.53517787 0.60063291
Xz 0.60000581 0.58685048 0.58270739
XH -0.00013 0 0
SSM 0.00086419 -0.0516726 0.01792552
stable not stable stable

Figure 6.49: SSM wing with endplate, h/c= 0.06 - 0.3, he/c = 0.06c
136

Table 6.43: SSM on rectangular-dihedral reverse wing NACA 6409 A/R = 1.25
Trimaran WIG model with endplate (0.06c) and NACA = 0012 Tail A/R = 5.

WITH ENDPLATE(0.06c)
ANALYSIS STATIC STABILITY MARGIN
(assume cog=0.3 c)

h/c h/c h/c


0.06 0.1 0.3
Cma -0.0138 -0.0119 -0.0095
Cla 0.046 0.0506 0.0316
Cmz 1.548 0.9738 0.274
Clz -5.1599 -3.3948 -0.9692
CmH 0.00000013 0 0
ClH -0.001 0 0
Cmat 0.0169 0.0169 0.0169
Clat 0.0675 0.0675 0.0675
Xa 0.601 0.53517787 0.60063291
Xz 0.60000581 0.58185048 0.58270739
XH -0.00013 0 0
Xat 0.04962963 0.04962963 0.04962963
SSM 0.05049382 0.00295702 0.06755515
stable stable stable

Figure 6.50: SSM wing with endplate and tail, h/c= 0.06 - 0.3 he/c = 0.06c
137

6.5 Longitudinal Dynamic Stability Trimaran WIG Model

These analysis were conducted to investigate the Dynamic stability during


take-off Trimaran WIG model with rectangular-dihedral reverse wing NACA 6409,
aspect ratio (AR) 1.25, with various ground clearance (h/c) = 0.06, 0.1, 0.2,0.3,
various angle of attack (AOA) 2 o,4 o,6 o,8 o
with endplate he/c = 0.06c, and tail
NACA 0012 A/R = 5. Figure 6.51 show Short Period Pitch Oscillation (SPPO)
perturbation every ground clearance (h/c) with constant angle of attack (2 deg).
Figure 6.52 show Short Period Pitch Oscillation (SPPO) perturbation every angle of
attack (AOA) with constant ground clearance (h/c=0.2). The magnitude of SPPO
perturbation decreases with increment of ground clearance (h/c), otherwise SPPO
perturbation increases with increment of angle of attack (AOA).

Figure 6.51: Short Period Pitch Oscillation (SPPO) Trimaran WIG every ground
clearance (h/c) during take-off
138

Figure 6.52: Short Period Pitch Oscillation (SPPO) Trimaran WIG every angle of
attack (AOA) during take-off

Figure 6.53 show Long Period Pitch Oscillation (Phugoid) perturbation every
ground clearance (h/c) with constant angle of attack (2 deg). Figure 6.54 show Long
Period Pitch Oscillation (Phugoid) perturbation every angle of attack (AOA) with
constant ground clearance (h/c=0.2). The magnitude of Phugoid perturbation
decreases with increment of ground clearance (h/c), otherwise Phugoid perturbation
increases with increment of angle of attack (AOA). From both figure we can see all
perturbation back into the equilibrium, regarding Tulapukara statement [0], an object
system is said to be dynamically stable if it eventually returns to the original
equilibrium position after being disturbed by a small disturbance. Also in these cases
the system while returning to the equilibrium position goes beyond the undisturbed
state towards the negative side and smaller than the original disturbance and it
(amplitude) decreases continually with every oscillation. And finally, the system
returns to the equilibrium position. So, dynamic stability Trimaran WIG in stable
condition and in damping oscillation. However we must check its dynamic quality to
be accepted for public transportation.
139

Figure 6.53: Long Period Pitch Oscillation (Phugoid) Trimaran WIG every ground
clearance (h/c) during take-off.

Figure 6.54: Long Period Pitch Oscillation (Phugoid) Trimaran WIG every angle
of attack (AOA) during take-off
140

6.6 Routh-Hurwitz Stability Criterion and Control Anticipation Parameter


(CAP)

Figure 6.55 shows comparison between Simulink Flightgear simulation


(detail simulation can bee see at Appendix H) and free running test, where from that
images are in a stable condition in visualization. Routh-Hurwitz Stability Criterion
and Control Anticipation Parameter (CAP) have been used to analyze dynamic
stability numerically Trimaran WIG during take-off. Detail calculation for Routh-
Hurwitz Stability Criterion can be used program at Appendix E, Short period
damping ratio should be bigger than 0.35 and less than 1.3, and Phugoid damping
ratio should be bigger than 0.4.

Figure 6.55: Comparison Trimaran WIG during take-off between Simulink


Flightgear Simulation and Free Running Test

Table 6.44 shows Routh-Hurwitz Stability Criterion for SPPO every ground
clearance (h/c), where all SPPO in stable condition. The magnitude of Eigen value
increases with increment of ground clearance (h/c).
141

Table 6.44: Routh-Hurwitz Stability Criterion for SPPO every ground clearance
(h/c)

Routh Hurwitz
h/c Equilibrium Eigen Value
criterion
-0.2989 s - 11.03
0.06 --------------------------------- -8.68e-002 +- 9.20e-001i accepted
s^3 + 9.063 s^2 + 2.399 s + 7.599

-0.2989 s - 11.03
0.1 --------------------------------- -7.50e-002 +- 9.22e-001i accepted
s^3 + 9.039 s^2 + 2.188 s + 7.599

-0.2989 s - 11.03
0.2 ------------------------------- -1.55e-001 +- 9.12e-001i accepted
s^3 + 9.198 s^2 + 3.606 s + 7.6

-0.2989 s - 11.03
0.3 --------------------------------- -1.21e-001 +- 9.17e-001i accepted
s^3 + 9.131 s^2 + 3.003 s + 7.599

Table 6.45 shows Control Anticipation Parameter (CAP) for SPPO every
ground clearance (h/c), where at h/c: 0.06-0.1 SPPO in not stable condition,
otherwise at h/c: 0.2-0.3 SPPO in stable condition. The magnitude of damping ratio
SPPO increases with increment of ground clearance (h/c), otherwise the magnitude
of natural frequency SPPO decreases with increment of ground clearance (h/c).

Table 6.45: Control Anticipation Parameter (CAP) for SPPO every ground
clearance (h/c).

damping CPT (damping CPT (natural


h/c Equilibrium natural freq
ratio check) freq check)
-0.2989 s - 11.03
---------------------------------
0.06 1.31E-01 9.25E-01 no accepted accepted
s^3 + 9.131 s^2 + 3.003 s +
7.599
-0.2989 s - 11.03
-------------------------------
0.1 1.67E-01 8.73E-01 no accepted accepted
s^3 + 9.198 s^2 + 3.606 s +
7.6
-0.2989 s - 11.03
---------------------------------
0.2 8.11E-01 6.56E-01 accepted accepted
s^3 + 9.039 s^2 + 2.188 s +
7.599
-0.2989 s - 11.03
---------------------------------
0.3 9.39E-01 4.67E-01 accepted accepted
s^3 + 9.063 s^2 + 2.399 s +
7.599
142

Table 6.46 shows Routh-Hurwitz Stability Criterion for Phugoid every


ground clearance (h/c), where all Phugoid in stable condition. The magnitude of
Eigen value decreases with increment of ground clearance (h/c).

Table 6.46: Routh-Hurwitz Stability Criterion for Phugoid every ground clearance
(h/c).

h/c Equilibrium Eigen Value Routh Hurwitz-criterion


11.03 s^2 + 3.207 s + 0.02067
------------------------------------------
0.06 -1.48e-002 +- 1.34e-001i accepted
0.8547 s^3 + 7.623 s^2 + 0.2402 s +
0.1387
11.03 s^2 + 2.458 s + 0.017
------------------------------------------
0.1 -1.48e-002 +- 1.25e-001i accepted
0.8547 s^3 + 7.623 s^2 + 0.2381 s +
0.1207
11.03 s^2 + 1.709 s + 0.01334
0.2 --------------------------------------- -1.48e-002 + 1.14e-001i accepted
0.8547 s^3 + 7.623 s^2 + 0.2358 s + 0.1
11.03 s^2 + 1.447 s + 0.01206
------------------------------------------
0.3 -1.48e-002 +- 1.04e-001i accepted
0.8547 s^3 + 7.623 s^2 + 0.234 s +
0.08391

Table 6.47 shows Control Anticipation Parameter (CAP) for Phugoid every
ground clearance (h/c), where all Phugoid in stable condition. The magnitude of
damping ratio Phugoid increases with increment of ground clearance (h/c), otherwise
the magnitude of natural frequency Phugoid decreases with increment of ground
clearance (h/c).
143

Table 6.47: Control Anticipation Parameter (CAP) for Phugoid every ground
clearance (h/c).

CPT
damping CPT (natural
h/c Equilibrium natural freq (damping
ratio freq check)
check)
11.03 s^2 + 3.207 s + 0.02067
------------------------------------------
0.06 1.09E-01 1.35E-01 accepted accepted
0.8547 s^3 + 7.623 s^2 + 0.2402 s +
0.1387
11.03 s^2 + 2.458 s + 0.017
------------------------------------------
0.1 1.17E-01 1.26E-01 accepted accepted
0.8547 s^3 + 7.623 s^2 + 0.2381 s +
0.1207
11.03 s^2 + 1.709 s + 0.01334
---------------------------------------
0.2 1.29E-01 1.15E-01 accepted accepted
0.8547 s^3 + 7.623 s^2 + 0.2358 s +
0.1
11.03 s^2 + 1.447 s + 0.01206
------------------------------------------
0.3 1.41E-01 1.05E-01 accepted accepted
0.8547 s^3 + 7.623 s^2 + 0.234 s +
0.08391

Table 6.48 shows Routh-Hurwitz Stability Criterion for SPPO every angle of
attack (AOA), where all SPPO in stable condition. The magnitude of Eigen value
increases with increment of every angle of attack (AOA).

Table 6.48: Routh-Hurwitz Stability Criterion for SPPO every angle of attack
(AOA)

h/c Equilibrium Eigen Value Routh Hurwitz-criterion

-0.2989 s - 11.03
2 --------------------------------- -3.92e-001 +- 8.37e-001i accepted
s^3 + 9.674 s^2 + 7.833 s + 7.604

-0.2989 s - 11.03
4 --------------------------------- -3.59e-001 +- 8.53e-001i accepted
s^3 + 9.606 s^2 + 7.229 s + 7.603

-0.2989 s - 11.03
6 --------------------------------- -3.25e-001 +- 8.66e-001i accepted
s^3 + 9.538 s^2 + 6.625 s + 7.603

-0.2989 s - 11.03
8 -------------------------------- -2.91e-001 +- 8.78e-001i accepted
s^3 + 9.47 s^2 + 6.022 s + 7.602
144

Table 6.49 shows Control Anticipation Parameter (CAP) for SPPO every
angle of attack (AOA), where at AOA: 2o - 6o SPPO in stable condition, otherwise at
AOA: 8o, SPPO in not stable condition. The magnitude of damping ratio SPPO
decreases with increment of every angle of attack (AOA), otherwise the magnitude
of natural frequency SPPO increases with increment of every angle of attack (AOA).

Table 6.49: Control Anticipation Parameter (CAP) for SPPO every angle of attack
(AOA)

damping CPT (damping CPT (natural freq


AOA Equilibrium natural freq
ratio check) check)
-0.2989 s - 11.03
---------------------------------
2 4.24E-01 9.25E-01 accepted accepted
s^3 + 9.674 s^2 + 7.833 s +
7.604
-0.2989 s - 11.03
---------------------------------
4 3.88E-01 9.50E-01 accepted accepted
s^3 + 9.606 s^2 + 7.229 s +
7.603
-0.2989 s - 11.03
---------------------------------
6 3.51E-01 9.76E-01 accepted accepted
s^3 + 9.538 s^2 + 6.625 s +
7.603
-0.2989 s - 11.03
--------------------------------
8 3.14E-01 9.87E-01 no accepted accepted
s^3 + 9.47 s^2 + 6.022 s +
7.602

Table 6.50 shows Routh-Hurwitz Stability Criterion for Phugoid every angle
of attack (AOA), where all Phugoid in stable condition. The magnitude of Eigen
value decreases with increment of every angle of attack (AOA).
145

Table 6.50: Routh-Hurwitz Stability Criterion for Phugoid every angle of attack
(AOA)

AOA Equilibrium Eigen Value Routh Hurwitz-criterion


11.03 s^2 + 8.451 s + 0.04632
-----------------------------------------
2 -1.48e-002 + 1.99e-001i accepted
0.8547 s^3 + 7.623 s^2 + 0.2585 s +
0.302
11.03 s^2 + 7.702 s + 0.04265
------------------------------------------
4 -1.48e-002 + 1.91e-001i accepted
0.8547 s^3 + 7.623 s^2 + 0.2559 s +
0.2787
11.03 s^2 + 6.953 s + 0.03899
------------------------------------------
6 -1.48e-002 +- 1.85e-001i accepted
0.8547 s^3 + 7.623 s^2 + 0.2539 s +
0.2607
11.03 s^2 + 6.204 s + 0.03532
------------------------------------------
8 -1.48e-002 + 1.78e-001i accepted
0.8547 s^3 + 7.623 s^2 + 0.2518 s +
0.2427

Table 6.51 shows Control Anticipation Parameter (CAP) for Phugoid every
angle of attack (AOA), where all Phugoid in stable condition. The magnitude of
damping ratio Phugoid increases with increment of ground clearance (h/c), otherwise
the magnitude of natural frequency Phugoid decreases with increment of ground
clearance (h/c).

Table 6.51: Control Anticipation Parameter (CAP) for Phugoid every angle of
attack (AOA)
damping CPT (damping CPT (natural
AOA Equilibrium natural freq
ratio check) freq check)
11.03 s^2 + 8.451 s + 0.04632
-----------------------------------------
2 7.41E-02 1.99E-01 accepted accepted
0.8547 s^3 + 7.623 s^2 + 0.2585 s +
0.302
11.03 s^2 + 7.702 s + 0.04265
------------------------------------------
4 7.72E-02 1.92E-01 accepted accepted
0.8547 s^3 + 7.623 s^2 + 0.2559 s +
0.2787
11.03 s^2 + 6.953 s + 0.03899
------------------------------------------
6 7.98E-02 1.85E-01 accepted accepted
0.8547 s^3 + 7.623 s^2 + 0.2539 s +
0.2607
11.03 s^2 + 6.204 s + 0.03532
------------------------------------------
8 8.27E-02 1.79E-01 accepted accepted
0.8547 s^3 + 7.623 s^2 + 0.2518 s +
0.2427
146

6.7 Dynamic Motion Trimaran WIG Model

Figure 6.56 shows comparison pitch amplitude between simulations and free
running test (see appendix G for detail experiments), where speed increase with
increment throttle, where from that images trend line simulation look similar with
experiment, but have difference outcome (-2.2% - 9.7%). Regarding that picture
pitch amplitude decrease with increment time, pitch angle looks larger at the
beginning of simulation, with a difference of about 0.5 deg from last simulation.
Very clear, the changing of phase during take-off really gives an effect in the pitch
motion Trimaran WIG model.

Figure 6.56: Pitch motion trimaran WIG model

Figure 6.57 shows comparison heave amplitude between simulations and free
running test (see appendix G for detail experiment) where speed increase with
increment throttle, where from that images trend line simulation look similar with
experiment, but have very significant difference outcome (-7.2% - 98.7%).
Regarding that picture heave amplitude increase with increment time, heave
amplitude looks lesser at the beginning of simulation, with a difference of about 1.5
cm. Very clear, the changing of phase during take-off really gives a effect in the
heave motion Trimaran WIG model.
147

Figure 6.57: Heave motion trimaran WIG model


CHAPTER 7

CONCLUSION AND RECOMMENDATION

7.1 Conclusion

Finally, the longitudinal stability and dynamic motion Trimaran WIG during
take-off was investigated. Regarding scope of work and methodology this research,
initial design trimaran WIG was made. A computational aerodynamic characteristic
of wing with endplate or without endplate was studied by VLM simulation.
Hydrodynamic WIG was investigated and was compared with commercial software.
Numerical was compared with experimental and commercial software. New
configuration Static Stability Margin (SSM) WIG trimaran model during take-off has
been created and validated. New configuration dynamic motion and longitudinal
dynamic stability Trimaran WIG during take-off was derived and solved. Detailed
conclusions from this research can be viewed as follows:

i. The traditional Vortex Lattice Method (VLM) has been modified by using
much complex Flat Wing Theory and then Wieselsberger Ground Effect
Theory has been used to calculate lift coefficient In Ground Effect (IGE).
VLM results were validated with the CFD calculation and experimental
data other researchers, where VLM results near experimental result. The
149

results show values of Lift Coefficient (CL) are more close to


experimental results for Angle of Attack (AOA) larger than 2 deg for
different Aspect Ratio (AR) and Ground Clearances. The magnitude of
drag coefficient increases with increment of aspect ratio, angle of attack,
and ground clearance. Drag coefficients of VLM at almost all angle of
attack more accurate than CFD results, where CFD far away to
experimental results. Researchers suggest this procedure can be followed
for initial estimation of Lift Coefficient for complex wing platform WIG.

ii. The problem of drag faced by WIG during take-off has been tried to be
solved by adopting sweep back step planing hull. Savitsky method has
been used to calculate planing hull and was enhanced with Hadler, Fox &
Blond, and clement formula. The calculation was validated using
commercial software .Results show difference about 8 7%.

iii. Trimaran WIG design has been made with new concept wing
Rectangular-Dihedral (ratio 50-50) Wing with aspect ratio (AR) 1.25 with
endplate (he/c) 0.06c, taper 0.8 with dihedral angle 13o at dihedral
reverse wing.

iv. The numerical aerodynamic Trimaran WIG model with and without
endplate has been calculated using modify VLM. The lift coefficient and
drag coefficient varied with angle of attack (0o, 2o, 4o, 6o, 8o, 10o), and
varied with ground clearances (from h/c = 0 until h/c = 0.15). The
magnitude of lift coefficient decreases with increment of ground
clearance, otherwise lift coefficient increase with increment angle of
attack. Lift coefficients wing with endplate increase around 14.5%. The
magnitude of drag coefficient increases with increment of ground
clearance, and drag coefficient increase with increment angle of attack.
Drag coefficients wing with endplate decrease around 25.17%.
150

v. The numerical Aerodynamic Trimaran WIG model was validated with


experimental at wind tunnel test LST-UTM. Lift coefficients of VLM at
0o - 8o angle of attack bigger than experimental results. The differences
that occur are 27.7% - 37.4% at 0o, 15.11% 15.9% at 2o, 7.55% - 8.9 %
at 4o, 2.3% - 4.55% at 6 o, 0.04% - 0.8% at 8 o. Drag coefficients of VLM
approximately at 0o angle of attack bigger than experimental results. But
at 2 o - 8 o angle of attack less than experimental results, difference case
occur at h/c= 0.15 almost drag coefficient VLM result like experimental
results. The differences that occur are 2.9% - 6.45% at 0 o, -0.3% -1.5%
at 2 o, -2.7% -17.3 % at 4 o, -16.3% 2.8% at 6 o, -4.08% - 8.9% at 8 o.
Results from experimental has been correction with blockage effect
correction and reduce experiment result approximately 0.01% - 4.7%.

vi. Hydrodynamic Trimaran WIG with step and without step has been
calculate and was validated with free running test. Trend line from free
running test really different with trend line from SM with step, but looks
similar with without step. Its mean no more effect stepped hull in this
model, because step in this model very small. So in this moment we
assume Trimaran WIG model without stepped hull. At 0 5 knot
numerical result (SM without step) near with experimental (free running
test) result (not planing hull phase), but from 6-24 knot experimental
result bigger than numerical result (SM without step). This phenomena
was happens because of aerodynamic drag (air drag wing and fuselage)
began to affect of experiment results, because numerical result (SM
without step) only for the hydrodynamic drag, and therefore need to be
added aerodynamic drag on the numerical results. From 24 28 knot
numerical results (SM without step) bigger than experimental results (free
running test). This moment was occurred because draft fuselage Trimaran
WIG was changed or all body Trimaran WIG was began fly (take off
condition).
151

vii. New configuration Static Stability Margin (SSM) WIG during take-off
has been created and validated with three criteria, first, the position
Aerodynamic centre in Pitch (ACP) should be located downstream of the
position Aerodynamic Centre in Height (ACH) , second, the position of
center of gravity (COG) of the craft should be located upstream of the
aerodynamic center of pitch (ACP) and third, the position Aerodynamic
Centre in Height (ACH) should be located upstream of Hydrodynamic
center in Pitch (HCP).

viii. Investigation the Static Stability Margin (SSM) during take-off


rectangular-dihedral reverse wing NACA 6409 Trimaran WIG has been
done. There is not a stable phase during take-off. To overcome instability
during the takeoff, tail added (NACA 0012, AR = 5) at the trimaran
design WIG. SSM become stable in all condition.

ix. New configuration dynamic motion WIG during take-off has been created
and solved. The classical aircraft motion has been modified by calculating
the aerodynamic force, hydrostatic and hydrodynamic forces, using a
small perturbation assumption the full equations of motion WIG during
takeoff are derived and solved.

x. Longitudinal dynamic stability and dynamic motion Trimaran WIG


during take-off have been calculated with several angle of attack and
ground clearance. The magnitude of Short Periodic Pitch Oscillation
(SPPO) perturbation decreases with increment of ground clearance (h/c),
otherwise SPPO perturbation increases with increment of angle of attack
(AOA). The magnitude of Phugoid perturbation decreases with increment
of ground clearance (h/c), otherwise Phugoid perturbation increases with
increment of angle of attack (AOA). Trimaran WIG in stable condition
and in damping oscillation.
152

xi. Routh-Hurwitz Stability Criterion and Control Anticipation Parameter


(CAP) have been used to analyze dynamic stability numerically Trimaran
WIG during take-off. A Trimaran WIG during take-off almost in stable
condition every angle of attack and ground clearance. There are not
accepted conditions at ground clearance 0.06 and 0.1, where those ground
clearances are take-off phase. Also not accepted condition at 8 deg AOA.
Finally, this Trimaran WIG design needs more improvement to be
accepted and used for public transportation.

xii. The changing of phase during take-off really gives very significant effect
in pitch and heaves amplitude.

7.2 Recommendation

In this thesis, there are still many shortcomings due to the limited time and
knowledge. In the future are expected to conduct research as follows:

i. Conducting the research with more variations of aspect ratio wing.


ii. Conducting the research with many variations rectangular-dihedral wing
composition.
iii. Conducting the research with many variations angle and position of swept
back step planing hull.
iv. Find solution to reduce spray coming out from planing hull.
v. Optimizing the procedure free running test to obtain aerodynamic
characteristics and hydrodynamic characteristics directly.
vi. Find other solution to solve unstable condition during take-off
153

vii. Need more improvement at Trimaran WIG design to be accepted and


used for public transportation.
viii. Create a new concept design marine transportation more effective, more
futuristic and environmental-friendly.
REFERENCES

A.I. Maskalik, K.V. Rozhdestvensky, and D.N. Sinitsyn (1998). A View of


The Present State of Research in Aero-and Hydrodynamics of
Ekranoplans. RTO AVT Symposium on Fluid Dynamics Problem of
Vehicles Operating Near or in The Air-Sea Interface, held in
Amsterdam, The Netherlands, 5-8 October 1998, and published in
RTO MP-15.
Arthur W. Carter (1961). Effect Ground Proximity On The Aerodynamic
Characteristics Of Aspect- Ratio-1 Airfoil With and Without
Endplates. NASA,TN-D-970.
B. Leon (2007). Wing in Ground Effect (WIG) Aircraft Aerodynamic. Project
assessment Scholl Of Mechanical engineering The University of
Adelaide, Australia.
Blount & Fox (1976). Small Craft Power Prediction, Marine Technology.
BPPT (2007). Inter Island Transportation. Annual report BPPT.
Couser P.R (1997). Calm Water Powering Prediction for High Speed
Catamaran. Fast '97,Sydney, Australia.
Collu, M., Patel, M. H. & Trarieux, F (2007). Unified Mathematical Model
For High Speed Hybrid (Air and Water-Borne) Vehicles. 2nd
International Conference on Marine Research and Transportation,
Ischia, Naples, Italy, 28-30.
D. Savitsky (1985). Chapter IV Planing Craft of Modern Ships and Craft.
Naval Engineers Journal.
D.Savitsky (1964). Hydrodynamic design of Planing Hull. SNAME.
D.Savitsky, M.F.Delorme, Raju Datla (2007). Inclusion Whisker Spray Drag
In Performance Prediction Method For High Speed Planing Hulls.
Marine Technology Vol 44 No.1.
Elliot GR (1937). Full Scale of Investigation Ground effect, NACA, Rep-
265.
E. P. Clement and J. D. Pope (1959). Graphs for predicting the resistance of
Large Stepless Planing hulls at High Speeds. DTMB Report 1318.
155

Eugene P.Clement and James D.Rope (1961). Stepless and Stepped Planing
Hull Graph for Performance Prediction and Design. DTMB,
Report-1490.
Eugene P.Clement (2006). A Configuration for a Stepped Planing Boat
Having Minimum Drag (Dynaplane Boat). This publication is available
on the web site of the International Hydrofoil Society:
http://www.foils.org.
Halloran M, O'Meara S (1999). Wing in Ground Effect Craft Review. DSTO
Aeronautical and Maritime Research Laboratory, Australia.
H. Ghassemi, M. Ghiasi (2007). A Combined Method for the Hydrodynamic
Characteristics of Planing Crafts. Ocean engineering Journal,
2007,doi:10.1016/j.oceaneng.
H.H. Chun and C.H. Chang (2002). Longitudinal Stability and Dynamic
Motions of a Small Passenger WIG Craft. Elsevier journal of ocean
engineering, vol. 29, p. 1145-1162.
H.H Chun and C.H. Chang (2003). Turbulence flow simulation for wings in
ground effect with two ground conditions: fixed and moving ground.
International Journal of maritime engineering, p.211-227.
Hiromichi Akimoto, Syozo Kubo and Motoki Tanaka (2004). Investigation
of the Canard Type Wing In Surface Effect Ship. The 2nd Asia Pacific
Workshop on Hydrodynamics, Bussan, Korea.
http://www.se-technology.com/ (1996). Wing in Ground Effect
Aerodynamics.
Irodov R.D (1970). Criteria of Longitudinal Stability of Ekranoplan.
Ucheniye Zapiski TSAGI, 1 (4), 63-74.
I. S. Ishak, S.Mat, T.M.Lazim, M.K.Muhammad, S.Mansor, M.Z.Awang
(2006). Estimation of Aerodynamic Characteristic of A Light
Aircraft. Jurnal Mekanikal No. 22, 64-74.
J.B.Hadler (1966). The Prediction of Power Performance of Planing Craft.
SNAME.
Kumar P.E (1968). An Experimental Investigation into the Aerodynamic
Characteristics of a Wing with and without Endplates in
Ground Effect. College of Aerodynamics, Cranfield, England, Rept.
Aero 201, 1968.
156

Kumar P.E (1972). Some Stability Problems of Ground Effect Vehicle in


Forward Motion. Aeronautical Quarterly, 18, 4152.
K.V. Rozhdestvensky (1998). Theoretical Analysis of Dynamics of A WIG
Vehicle in Extreme Ground Effect. RTO AVT Symposium on Fluid
Dynamics Problem of Vehicles Operating Near or in The Air-Sea
Interface, held in Amsterdam, Netherlands, 5 - 8 October 1998, and
published in RTO MP-15.
K.V. Rozhdestvensky (2000). Aerodynamics of a Lifting System in Extreme
Ground Effect. Springer, Moscow.
K.V. Rozhdestvensky (2006). Wing-in-ground effect vehicles, Progress in
Aerospace Sciences. 42, (3), 211-283.
Kwang Hyo Jung, Ho Hwan Chun, Hee Jun Kim (2008). Experimental
Investigation of Wing-in-Ground Effect with a NACA 6409 Section.
Springer Jasnoe.
Kyoungwoo P. and J. Lee (2009). Optimal Design of 2-Dimensional Wings
in Ground Effect Using Multi-Objective Genetic Algorithm.
Manuscript Ocean Engineering.
Maurizio Collu, Minoo H. Patel and F. Trarieux (2010). The Longitudinal
Static Stability of an Aerodynamically Alleviated Marine Vehicle, a
Mathematical Model. Proc. R. Soc. A 2010 466, 1055-1075 first
published online 2 December 2009, doi: 10.1098/rspa.2009,0459.
M. Adi, P.Agoes, S. Noverdo, S.Ike, Mobasser, J. Saeed (2010).
Investigation on the Stability of a Trimaran wing in Ground Effect
(WIG) Craft with Endplate. International Conference on Marine
Technology (MARTEC 2010). The Department of Naval
Architecture and Marine Engineering (NAME), Bangladesh
University of Engineering and Technology (BUET), Dhaka, 11-12
December.
Nicola de Divitiis (2009). Performance and Stability of Winged in Ground
Effect. ArXiv: 0912.3355v1.
N. Kornev (2003). Complex Numerical Modeling of Dynamics and Crashes
of Wing in Ground Vehicles. 41st Aerospace Sciences Meeting and
Exhibit, Reno, Nevada.
157

Paul E. Hemke (1927). Drag Of Wing With End Plates. NACA, Rep 267.
Prandtl, Ludwig (1920). Theory of Lifting Surface. NACA, TN 09.
Rae, William H. Jr., Pope, Alan (1984). Low-Speed Wind Tunnel Testing. John
Wiley & Sons.
Richard J. Margason and John E. Lamar (1971). Vortex Lattice Fotran
Program for Estimating Subsonic Aerodynamic Characteristic of
Complex Planforms. NASA TN D-6142.
R. Savander Brant & Hyung Rhee Shin (2003). Steady Planing
Hydrodynamics: Comparison of Numerical and Experimental
Results. Presentation outline Fluent Users Group Manchester, NH.
Rozhdestvensky K.V (1996). Ekranoplans The GEMs of Fast Water
Transport. Trans ImarE, Vol 109, Part 1, pp 47-74.
Saaed J., S. Noverdo (2010). Effect Ground Effect for Environmental. 3rd
International Graduate Conference on Engineering, Science and
Humanities (IGCESH 2010) UTM, Malaysia, November 2 4.
Saunders H.E (1957). Hydrodynamics in Ship Design. Vol. 1. The Society of
Naval Architects and Marine Engineers, New York.
S.C Rhodes and A.T Sayers (2009). Experimental Investigation: Stability
criteria of an Uncambered Airfoil in Ground Effect. R & D Journal
of the South African Institution of Mechanical Engineering 2009, 25.
S. Noverdo, Mobasser (2010). Effect of Stepped Hull on Wing in Ground
Effect (WIG) Craft during Takeoff. 3rd International Graduate
Conference on Engineering, Science, and Humanities (IGCESH
2010) UTM, Malaysia, November 2 4.
Staufenbiel RW and Schlichting UJ (1988). Stability of Airplanes in Ground
Effect. Journal of Aircraft, 25, (4), 289-294.
Staufenbiel R and Kleineidam G (1980). Longitudinal Motion of Low-
Flying Vehicles in Nonlinear Flowfields. Proceedings of the
Congress of the International Council of the Aeronautical Sciences,
Munich, 293308.
Suhaimi (2008). Low Speed Wind Tunnel. Specification LST UTM Facility.
Takinaci A.C, Atlar M, Korkut E (2002). A Practical Surface Panel Method
to Predict Velocity Distribution around a Three-Dimensional
158

Hydrofoil Including Boundary Layer Effects. Elsevier journal of


ocean engineering, PII: S00 29 -8018(02)00015-X.
Todd E. Whalen (2002). Optimal Deadrise Hull Analysis and Design Space
Study of Naval Special Warfare High Speed Planing Boats. Masters
thesis Department Master of Science in Naval Architecture and
Marine Engineering and Master of Science in Civil and
Environmental Engineering, Massachusetts Institute of Technology.
Tulapurkara E.G (2008). Flight Dynamics-I (Performance Analysis).
Department. of Aerospace Engineering, IIT Madras, Chennai-
600036, India.
V.I. Korolyov (1998). Longitudinal Stability of Ekranoplans and Hydrofoils
Ships. RTO AVT Symposium on Fluid Dynamics Problem of
Vehicles Operating Near or in The Air-Sea Interface, held in
Amsterdam, The Netherlands, 5-8 October 1998, and published in
RTO MP-15.
Wieselsberger C (1921). Wing Resistance near the Ground, NACA, TM-
77.
Zhang, Zerihan, Ruhrmann (2002). Tip vortices generated by wing in
ground effect. Proceedings of the First International Symposium on
Applications of Laser Techniques to Fluid Mechanics, Portugal.
Zhukov V.I (1974). Some matters of Longitudinal Stability of Ekranoplans.
Trudy TSAGI.
159

APPENDIX A

TRIMARAN WIG DESIGN

A. 1 Preliminary Design

Table A.1 Preliminary Design

PRELIMINERY DESIGN HULL


Displacement 4.7 kg
LPP 1200 mm
Lwl (hull +wing) 1082.8 mm
Beam wl 858.4 mm
Volume 4543826.4 mm^3
Draft to Baseline 60 mm
Immersed depth 58.9 mm
WSA 303388.2 mm^2
Max cross sect area 9613.8 mm^2
Waterplane area 138566.7 mm^2
160

Figure A.1 Front View Trimaran WIG

Figure A.2 Side View Trimaran WIG

Figure A.3 Top View Trimaran WIG


161

Figure A.3 Lines Plane

Figure A.3 Wing section


162

A.2 Hydrostatic Calculation

Table A.1 Hydrostatics Calculation

HYDROSTATICS
Draft Amidsh. cm 6.508
Displacement kg 5.4
Heel to Starboard degrees 0
Draft at FP cm 6.141
Draft at AP cm 6.875
Draft at LCF cm 6.522
Trim (+ve by stern) cm 0.735
Trim (+ve by stern) deg 2.776
WL Length cm 112.5
WL Beam cm 85.69
Wetted Area cm^2 3637
Waterpl. Area cm^2 1423
Prismatic Coeff. 0.445
Block Coeff. 0.391
Midship Area Coeff. 0.879
Waterpl. Area Coeff. 0.679
LCB from Amidsh. (+ve fwd) cm 0
LCF from Amidsh. (+ve fwd) cm -1.81
KB cm 4.332
KG cm 6.031
BMt cm 77.68
BML cm 200.5
GMt cm 75.98
GML cm 198.8
KMt cm 82.01
KML cm 204.9
Immersion (TPc) tonne/cm 0.001
MTc tonne.m 0
RM at 1deg = GMt.Disp.sin(1)
kg.cm 7.16
Max deck inclination deg 0.5
Trim angle (+ve by stern) deg 0.5
163

APPENDIX B

AIRFOIL ANALYSIS

B.1 Introduction

These analyses have been done for choosing suitable foil for TRIMARAN
WIG using 2D potential flow. These analyses divided into two conditions, out
ground condition and in ground condition

B.2 Out Ground condition

In this simulation using NACA 6409 with angle of attack (AOA) 0o-6o out
ground condition.
164

Figure B.1 Flow field out ground condition NACA 6409

Figure B.2 Velocity distribution out ground condition NACA 6409


165

Figure B.3 Pressure Coefficient distribution out ground condition NACA 6409

Figure B.4 Boundary layer out ground condition NACA 6409


166

Figure B.5 Initial Lift Coefficient out ground condition NACA 6409
167

B.2 In Ground condition

In this simulation using NACA 6409 with angle of attack (AOA) 0o-6o in
ground condition h/c= 0.1c.

Figure B.6 Flow Field in ground condition NACA 6409


168

Figure B.7 Velocity distribution in ground condition NACA 6409

Figure B.8 Pressure Coefficient distribution in ground condition NACA 6409

Figure B.9 Boundary layer in ground condition NACA 6409


169

Figure B.10 Initial Lift Coefficient in ground condition NACA 6409


170

APPENDIX C

AERO-HYDRO DERIVATIVES

The hydrodynamic and aerodynamic derivative component at the first and


second order Taylor expansion equation motion will be discussed in this chapter.
The hydrodynamic characteristic will concentrate for the wig planing hull form
instead the aerodynamic characteristic will concentrate for whole wig craft.

C.1 Hydrodynamics derivative

The force equation at equilibrium state for surge and heave can write as:

X = FLWsin + FSWsin - DWcos (C.1)


Z = -FLWcos FSWcos - DWsin (C.2)
171

Where FLW and FSW represent the hydrodynamic force and static buoyancy
force respectively which both perpendicular to the V direction, DW represent the
drag force due to resistance force parallel to the V direction.
The terms of force in equation (C.1) and (C.2) usually represent by
coefficient as the expression below:

CLW = FLW/(1/2wV2Sah)
CSW = FSW/(1/2wV2Sah)
CDW= DW/ (1/2wV2Sah)

Where CLW, CSW, CDW are coefficient of hydrodynamic force, static


buoyancy force and drag force respectively. While w is water density, Sah is planing
hull form reference area, V is the magnitude of flow velocity.

The equation (C.1) and (C.2) can express as form below:

X = 1/2wV2Sah (CLWsin + CSWsin - CDWcos) (C.3)


Z = -1/2wV2Sah (CLWcos + CSWcos - CDWsin) . (C.4)

From the literature study by apply small perturbation theory; the following
approximation can be made:

sin (C.5)
cos 1 (C.6)

By apply this approximation, equation (C.3) and (C.4) become:


172

X = 1/2wV2Sah (CLW + CSW - CDW) (C.7)


Z = -1/2wV2Sah (CLW + CSW - CDW) (C.8)

Relationship also can be expressed as:

V= (U0+u) 2 + v2 +w2 (C.9)


 
= = cos 1 (C.10)
 
 
= = sin 0 (C.11)
 
 

= 0 (C.12)
 

=  (C.13)
 

In order to derive Xu, the X have to assume is a function of V and ,


X=f (V, ), then apply the Chain rule. The Xu derivative with superscript h can be
expressed as below:

      
X uh = = =   (C.14)
      

By standard approximation

  Sw 


= 0,  0, 
 0, 
 0, 
0 (C.15)

X uh can be made as below:


173


X uh =  = -wVSahCDw (C.16)

Also can be expressed as (0):

X uh = h g ( B h ) 2 (C.17)

Where B h beam of planning hull

In order to derive Zw, the Z also have to assume is a function of V and ,


Z=f (V, ), then apply the Chain rule. The Zw derivative as below:

     
Z wh = =   (C.18)
      
 
Z wh = wVSah ( + + CDW) (C.19)
 

Following is the moment about Y-axis means pitching moment derivative.


From the figure illustration, assume the distance between hydrodynamic centers to
center of gravity is lhx along X-axis and lhz along Z-axis. While the distance between
buoyancy center to center of gravity is assume lbx along X-axis and lbz along Z-
axis. The moment about the center gravity can express as below:

M wh = X wh (lhz + l bz ) + Z wh (lhx + l hz ) (C.20)

Z wh already found at previous stage, X wh can be expressed like previous method:

     
X wh =    

 

 
(C.21)
174


X wh = wSah (CLw + CSw - 
) (C.22)

Other hydrodynamic derivatives can be found with previous method. All


hydrodynamic derivatives have summarized in this table;

X uh = h g ( B h ) 2 Z uh = h g ( B h ) 2 M uh = h g ( B h )3
X qh = h g ( B h )3 Z qh = h g ( B h )3 M qh = h g ( B h ) 4
X h. = h g ( B h )3 g / B h Z .h = h g ( B h )3 g / B h M h. = h g ( B h ) 4 g / B h
u u u

h 3 h 3
h h
X . = g(B ) g/B h h h
Z . = g(B ) g/B h
M h. = h g ( B h ) 4 g / B h
w w w

X h. = h g ( B h ) 4 g / B h Z h. = h g ( B h ) 4 g / B h M h. = h g ( B h ) 5 g / B h
q q q
h 3 h 3
h h
X .. = g ( B ) h h
Z .. = g ( B ) X .. = h g ( B h ) 4
h
u u u

X ..h = h g ( B h )3 Z ..h = h g ( B h ) 3 X ..h = h g ( B h ) 4


w w w
h 4 h 4
h h
X .. = g ( B ) h h
Z .. = g ( B ) X ..h = h g ( B h )5
w w w

C.2 Aerodynamics derivative

The aerodynamic derivative concepts are similar to the hydrodynamic


derivative. The different between two derivatives are the free body diagram and the
reference parameter applying in derivative such as the reference area, external force
acting on the wig craft.

The force equation at equilibrium state for surge and heave can write as:

X = FLAsin DAcos (C.23)


Z = -FLAcos DAsin (C.24)
175

Where FLA represent the aerodynamic force which perpendicular to the V


direction, DA represent the drag force due to resistance force parallel to the V
direction.

The terms of force in equation (C.23) and (C.24) usually represent by


coefficient as the expression below:

CLA = FLA/ (1/2AV2Saw) (C.25)


2
CDA= DA/ (1/2AV Saw) (C.26)

Where CLA, CDA are coefficient of aerodynamic lift force and drag force
respectively. While A is air density, Saw is wig crafts main wing reference area, V
is the magnitude of flow velocity similar with the V in hydrodynamic derivative.

Although the concept of derivative are similar for both characteristic but the
Zq derivative is not exist in hydrodynamic. In order to Zq, the wig craft pitching
rate,q impart a downward velocity to the horizontal tail of magnitude WT, express as
WT = qlT, where lT is the tail arm. The downward velocity at the tail is assuming lead
to an additional force in Z-axis. This force, ZT express as below:


ZT =  WT

 
= -1/2ASaw ( + ! ) qlT (C.27)


Using equation (C.27) Z qa can be expressed as:


176

Z T C T
Z qa = T
= A S w ( LA + C DA )lT (C.29)
q

All derivative of aerodynamic can be found with previous method, and and
the results can be viewed as follows:

CD CL CM
X ha = Z ha = CL M ha =
(h / c) (h / c) (h / c)
CD C M ua = neglegible (0)
X ua = 2CD V Zua = 2CL + V L
V V
CD C CM
X wa = CL Z wa = CD L M wa =

a
C S L 2
X q = neglegible(0) C S L
Z ..a = LT T T M ..a = LT T T
w S c q Sc 2
2
X a. = neglegible(0) CLT ST LT
w M ..a =
w Sc 2
177

APPENDIX D

VORTEX LATTICE METHOD SOFTWARE

% Modified for
% DEVELOPMENT WING IN GROUND EFFECT TWO SETTER
% UNIVERSITY TECHNOLOGY MALAYSIA
% MODIFIED BY N.S

clc
fprintf('WIG 2 SETTER UTM\n')
clear all
close all

%Parse NACAdata.txt into airfoilstruct for use in determining


profile drag
airfoilstruct = readNACAairfoildata;
%Perform regression on airfoilstruct data at multiple Re
airfoildata = nacaRegression(airfoilstruct);

%Initialize GUI
InitializeGUI(airfoildata)

function [finaldata] = readNACAairfoildata


% parse the text file NACAdata.txt into a structure

fid = fopen('NACAdata.txt','r');

data = textscan(fid, '%s %n %n %n %n %n');

fclose(fid);

%Initialize finaldata
finaldata(1).airfoil = data{1}{1};
178

finaldata(1).Re = data{2}(1);
finaldata(1).a0 = data{3}(1);
finaldata(1).a1 = data{4}(1);
finaldata(1).a2 = data{5}(1);
finaldata(1).alphastall = data{6}(1);

%Iterate through all entries in data

for i = 2:length(data{1})
updated = 0;
for j = 1:length(finaldata)
%if we have already read in data about a certain airfoil,
append
%new data
if strcmp(finaldata(j).airfoil,data{1}{i})
finaldata(j).Re = [finaldata(j).Re; data{2}(i)];
finaldata(j).a0 = [finaldata(j).a0; data{3}(i)];
finaldata(j).a1 = [finaldata(j).a1; data{4}(i)];
finaldata(j).a2 = [finaldata(j).a2; data{5}(i)];
finaldata(j).alphastall = [finaldata(j).alphastall;
data{6}(i)];
updated = 1;
break
end
end
%if we have not already read in data about a certain airfoil,
create
%new entry
if updated == 0
finaldata(end+1).airfoil = data{1}{i};
finaldata(end).Re = data{2}(i);
finaldata(end).a0 = data{3}(i);
finaldata(end).a1 = data{4}(i);
finaldata(end).a2 = data{5}(i);
finaldata(end).alphastall = data{6}(i);
end
end

function [CL,CM, u_U, v_U, w_U, l_t, l_s] = LiftCoeff(gamma, panel,


geo, Fu_bar, Fv_bar, Fw_bar)

%Determine the lift coefficient given the vortex strengths


%and the influence coefficients from the vortex lattice method

u_U.c = 1/(4*pi)*Fu_bar*[gamma.c]'; %Eqn 25 in NASA paper (NASA TN


D6142)
v_U.c = 1/(4*pi)*Fv_bar*[gamma.c]'; %Eqn 23 in NASA paper (NASA
TN D6142)
w_U.c = 1/(4*pi)*Fw_bar*[gamma.c]';

u_U.alpha = 1/(4*pi)*Fu_bar*[gamma.alpha]';
v_U.alpha = 1/(4*pi)*Fv_bar*[gamma.alpha]';
w_U.alpha = 1/(4*pi)*Fw_bar*[gamma.alpha]';
179

for i = 1:geo.ns
for j = 1:geo.nc
if i == geo.ns %wingtip
DeltaGamma(i,j).c = gamma(i,j).c;
DeltaGamma(i,j).alpha = gamma(i,j).alpha;
elseif j == 1 %Leading edge
DeltaGamma(i,j).c = gamma(i,j).c-gamma(i+1,j).c;
DeltaGamma(i,j).alpha = gamma(i,j).alpha-
gamma(i+1,j).alpha;
else
DeltaGamma(i,j).c = DeltaGamma(i,j-1).c + gamma(i,j).c -
gamma(i+1,j).c;
DeltaGamma(i,j).alpha = DeltaGamma(i,j-1).alpha +
gamma(i,j).alpha - gamma(i+1,j).alpha;
end
end
end

l_t.c = 2/geo.S*[DeltaGamma.c]'.*[panel.cc]'.*[v_U.c]; %Eqn 24 in


NASA paper (NASA TN D6142)

l_t.alpha = 2/geo.S*[DeltaGamma.alpha]'.*[panel.cc]'.*[v_U.alpha];
%Eqn 24 in NASA paper (NASA TN D6142)

l_s.c = 2/geo.S*[gamma.c]'*2.*[panel.s]'.*((ones(size(u_U))-
[u_U.c])+[v_U.c].*tan([panel.sweep]'))*cos(geo.dih); %Eqn 28 in
NASA paper
l_s.alpha = 2/geo.S*[gamma.alpha]'*2.*[panel.s]'.*((ones(size(u_U))-
[u_U.alpha])+[v_U.alpha].*tan([panel.sweep]'))*cos(geo.dih); %Eqn
28 in -NASA paper

% reshape([l_t.c],geo.ns,geo.nc)
% reshape([l_t.alpha],geo.ns,geo.nc)
% reshape([l_s.c],geo.ns,geo.nc)
% reshape(((ones(size(u_U))-
[u_U.alpha])+[v_U.alpha].*tan([panel.sweep]')),geo.ns,geo.nc)
% reshape([panel.s],geo.ns,geo.nc)
% reshape([l_s.alpha],geo.ns,geo.nc)
% reshape([u_U.alpha],geo.ns,geo.nc)
% reshape([v_U.alpha],geo.ns,geo.nc)
% reshape([gamma.alpha],geo.ns,geo.nc)
% sum([gamma.c])
% geo.S

CL.c = 2*(sum([l_t.c])+sum([l_s.c]));
CL.alpha = 2*(sum([l_t.alpha])+sum([l_s.alpha]));
%asummsi CM 0.25c
CM.c = 2*((sum([l_t.c])*0.25)+(sum([l_s.c])*0.25)); %equation 29
nasa TN D6142
CM.alpha = 2*((sum([l_t.alpha])*0.25)+(sum([l_s.alpha])*0.25));

function [C,T,A]=naca4_3d(airfoil,y,nc)
180

if length(airfoil) == 4
m = str2num(airfoil(1))/100;
p = str2num(airfoil(2))/10;
t = str2num(airfoil(3:4))/100;

elseif length(airfoil) == 5
switch airfoil(1:3)
case '210'
m = 0.0580;
k1 = 361.4;
case '220'
m = 0.1260;
k1 = 51.64;
case '230'
m = 0.2025;
k1 = 15.957;
case '240'
m = 0.2900;
k1 = 6.643;
case '250'
m = 0.3910;
k1 = 3.230;
otherwise
end
t = str2num(airfoil(4:5))/100;
end

%% Constants
% nc % number of line segments describing top surface
% (= number of line segments describing bottom surface)
npc = nc+1; % number of points along mean camber line
nps = 2*nc+1; % number of points along surface
dx=1/nc; % x-direction increment (along chord)

%% Fill coordinate matrices with y coordinate

C = zeros(npc,3); % mean camber


C(:,2) = y;
T = zeros(nps,3); %skin
T(:,2) = y;

%% Mean Camber Line and Slope


% dzcdx is the derivative of the mean camber line
% theta is the local angle of the mean camber line
% x=1:-dx:0; % order x locations starting at trailing edge, x=1

x=0:dx:1; % order x locations starting at leading edge, x=0


if length(airfoil) == 4
for i = 1:nc+1
if x(i)<p % leading edge
zc(i) = m/(p^2)*(2*p*x(i)-x(i)^2); %Eqn 6.4 Abbot
dzcdx(i) = m/p^2*(2*p-2*x(i));
theta(i) = atan(dzcdx(i));
else % trailing edge
181

zc(i) = m/(1-p)^2*((1-2*p)+2*p*x(i)-x(i)^2); %Eqn 6.4


Abbot
dzcdx(i) = m/(1-p)^2*(2*p-2*x(i));
theta(i) = atan(dzcdx(i));
end
end
elseif length(airfoil) == 5
for i = 1:nc+1
if x(i)<m % leading edge
zc(i) = 1/6*k1*(x(i)^3-3*m*x(i)^2+m^2*(3-m)*x(i)); %Eqn
6.6 Abbot
dzcdx(i) = 1/6*k1*(3*x(i)^2-6*m*x(i)+m^2*(3-m));
theta(i) = atan(dzcdx(i));
else % trailing edge
zc(i) = 1/6*k1*m^3*(1-x(i)); %Eqn 6.6 Abbot
dzcdx(i) = -1/6*k1*m^3;
theta(i) = atan(dzcdx(i));
end
end
end

C(:,1)=-x; % output mean camber line coordinates; negative x due to


axis convention
C(:,3)=-zc; %negative z due to axis convention

%% Thickness distribution is the same for both 4- and 5-digit


airfoils

for i=1:nc+1
zt(i) = t/0.2*[0.2969*x(i)^0.5-0.1260*x(i)-
0.3516*x(i)^2+0.2843*x(i)^3-0.1015*x(i)^4]; %Eqn 6.2 Abbott
end

%% Area
A=0;
for i=1:nc
da=dx*(zt(i)+zt(i+1)); %Take the average of zt(i) and zt(i+1)
and double it
A=A+da;
end
A;

%% Upper skin

for i=1:nc+1
xu(i) = x(i)-zt(i)*sin(theta(i)); %Eqn 6.1 Abbott
zu(i) = zc(i)+zt(i)*cos(theta(i)); %Eqn 6.1 Abbott
T(i,1) = -xu(i); %negative x due to sign convention
T(i,3) = -zu(i); %negative z due to sign convention
end

%% lower skin
% exclude point nc+1 at leading edge
% because it is included as the last point of the upper surface
182

for i=1:nc
xl(i) = x(i)+zt(i)*sin(theta(i)); %Eqn 6.1 Abbott
zl(i) = zc(i)-zt(i)*cos(theta(i)); %Eqn 6.1 Abbott
end

% reverse order so LE to TE
for i=nc+2:nps
T(i,1) = -xl(2*(nc+1)-i);%negative x due to sign convention
T(i,3) = -zl(2*(nc+1)-i);%negative z due to sign convention
end

%figure
%plot(xu,zu,'o')
%hold on
%plot(xl,zl,'or')
%axis equal

function PARAM(H0,SIG,B)
global A0 A1 A3
E=(H0-1.0)/(H0+1.0);
if(abs(SIG-pi/4.0)<=0.0001)
AL3=0.0;
else
E2=E*E;
C=4.0*SIG*(1.0-E2)/pi+E2;
D=-C/(C+3.0);
D2=sqrt(D*D-(C-1.0)/(C+3.0));
AL3=D+D2;
end
AL1=E*(AL3+1.0);
A0=B/(1.0+AL1+AL3);
A1=A0*AL1;
A3=A0*AL3;

function [y_cp] = SpanLoading(l_s, l_t, CL, geo, panel)


%Determine center of pressure and spanwise loading
%due to bound vortices and trailing vortices.

BV = [panel.BV];
BV1 = [panel.BV1];
s = [panel.s]';

y_cp.alpha = sum([l_s.alpha].*BV(2,:)' +
[l_t.alpha].*BV1(2,:)')/(0.5*CL.alpha*0.5*geo.b); %Eqn 35 in NASA
paper
cl.alpha = ([l_s.alpha] +
[l_t.alpha])*geo.S./(CL.alpha*2*s*cos(geo.dih)*geo.c_av); %Eqn 37
in NASA paper

if CL.c ~= 0 %Prevent divide by zero


y_cp.c = sum([l_s.c].*BV(2,:)' +
[l_t.c].*BV1(2,:)')/(0.5*CL.c*0.5*geo.b); %Eqn 35 in NASA paper
cl.c = ([l_s.c] + [l_t.c])*geo.S./(CL.c*2*s*cos(geo.dih)*geo.c_av);
%Eqn 37 in NASA paper
else
y_cp.c = 0;
183

cl.c = zeros(size(cl.alpha));
end

function [dens, a, visc] = StandardAtmosphere(alt)


%Read in altitude and output viscosity, density, and speed of sound
%Units:
%altitude as a string in ft
%density in kg/m^3
%speed of sound, a, in m/s

%Convert altitude to number in meters


alt = str2num(alt)*0.3048;

g = 9.80665;
R = 287.1;
gamma = 1.4;
if alt < 11000
T1 = 288.16; %From Anderson, Introduction to Flight, 5 ed, pg
109
laps = -6.5e-3;
T = T1 + laps*(alt);
dens1 = 1.2250;
dens = (T/T1)^-(g/(laps*R)+1)*dens1;
elseif alt >= 11000 & alt < 25000
T1 = 216.66;
T = T1;
dens1 = 0.3640;
dens = exp(-g/(R*T)*(alt-11000))*dens1;
else
T1 = 216.66;
laps = 3e-3;
T = T1 + laps*(alt);
dens1 = 0.11;
dens = (T/T1)^-(g/(laps*R)+1)*dens1;
end
a = sqrt(gamma*R*T);
visc = 1.458e-6*T^1.5/(T+110.1); %Bertin, Aerodynamics for
Engineers, 4 ed, pg 5
function [gamma, Fu_bar, Fv_bar, Fw_bar]=VortexStrength(panel,phi)

% Implement Vortex Lattice Method


%Input: panel geometry
%Output vortex strength (gamma)
%Influence coefficients (Fx_bar)

ns = size(panel,1);
nc = size(panel,2);

%Preallocate memory for large variables where first dimension points


to controls points; second dimension points to horseshoe vortices;
% third dimension points to three dimensional coordinates

r0 = zeros(ns*nc,2*ns*nc,3);
r1 = zeros(ns*nc,2*ns*nc,3);
184

r2 = zeros(ns*nc,2*ns*nc,3);
omega_bound = zeros(ns*nc,2*ns*nc,3);
omega_a_inf = zeros(ns*nc,2*ns*nc,3);
omega_b_inf = zeros(ns*nc,2*ns*nc,3);
Fu = zeros(ns*nc,2*ns*nc);
Fv = zeros(ns*nc,2*ns*nc);
Fw = zeros(ns*nc,2*ns*nc);

%Determine influence coefficients

%Iterate through all control points


for i = 1:ns
for j = 1:nc
n=(j-1)*ns+i;
%Iterate through all bound vortices
for k = 1:ns
for p=1:nc
m=(p-1)*ns+k;

%Code derived from Bertin, Aerodynamics for


Engineers, ed.
%4, pg 263-266
r0(n,m,:)=panel(k,p).BV2-panel(k,p).BV1;
r1(n,m,:)=panel(i,j).CP-panel(k,p).BV1;
r2(n,m,:)=panel(i,j).CP-panel(k,p).BV2;

Fac1=shiftdim(cross(r1(n,m,:),r2(n,m,:)))/(norm(shiftdim(cross(r1(n,
m,:),r2(n,m,:)))))^2;

Fac2=dot(r0(n,m,:),(r1(n,m,:)/norm(shiftdim(r1(n,m,:)))))-
dot(r0(n,m,:),(r2(n,m,:)/norm(shiftdim(r2(n,m,:)))));

Fac1a(1,1)=0;
Fac1a(1,2)=(panel(i,j).CP(3)-
panel(k,p).BV1(3))/((panel(i,j).CP(3)-
panel(k,p).BV1(3))^2+(panel(k,p).BV1(2)-panel(i,j).CP(2))^2);
Fac1a(1,3)=(panel(k,p).BV1(2)-
panel(i,j).CP(2))/((panel(i,j).CP(3)-
panel(k,p).BV1(3))^2+(panel(k,p).BV1(2)-panel(i,j).CP(2))^2);
Fac2a=-1+((panel(i,j).CP(1)-
panel(k,p).BV1(1))/norm(shiftdim(r1(n,m,:)))); %'-1' due to change
in axes

Fac1b(1,1)=0;
Fac1b(1,2)=-(panel(i,j).CP(3)-
panel(k,p).BV2(3))/((panel(i,j).CP(3)-
panel(k,p).BV2(3))^2+(panel(k,p).BV2(2)-panel(i,j).CP(2))^2);
Fac1b(1,3)=-(panel(k,p).BV2(2)-
panel(i,j).CP(2))/((panel(i,j).CP(3)-
panel(k,p).BV2(3))^2+(panel(k,p).BV2(2)-panel(i,j).CP(2))^2);
Fac2b=(-1+((panel(i,j).CP(1)-
panel(k,p).BV2(1))/norm(shiftdim(r2(n,m,:)))));

omega_bound(n,m,:)=Fac1*Fac2;
omega_a_inf(n,m,:)=Fac1a*Fac2a;
omega_b_inf(n,m,:)=Fac1b*Fac2b;
185

Fu(n,m)=omega_bound(n,m,1)+omega_a_inf(n,m,1)+omega_b_inf(n,m,1);

Fv(n,m)=omega_bound(n,m,2)+omega_a_inf(n,m,2)+omega_b_inf(n,m,2);

Fw(n,m)=omega_bound(n,m,3)+omega_a_inf(n,m,3)+omega_b_inf(n,m,3);
%Performing identical operations on opposite wing
r0(n,m+ns*nc,:)=[panel(k,p).BV1(1) -
panel(k,p).BV1(2) panel(k,p).BV1(3)]'-[panel(k,p).BV2(1) -
panel(k,p).BV2(2) panel(k,p).BV2(3)]';
r1(n,m+ns*nc,:)=panel(i,j).CP-[panel(k,p).BV2(1) -
panel(k,p).BV2(2) panel(k,p).BV2(3)]';
r2(n,m+ns*nc,:)=panel(i,j).CP-[panel(k,p).BV1(1) -
panel(k,p).BV1(2) panel(k,p).BV1(3)]';

Fac1=shiftdim(cross(r1(n,m+ns*nc,:),r2(n,m+ns*nc,:)))/(norm(shiftdim
(cross(r1(n,m+ns*nc,:),r2(n,m+ns*nc,:)))))^2;

Fac2=dot(r0(n,m+ns*nc,:),(r1(n,m+ns*nc,:)/norm(shiftdim(r1(n,m+ns*nc
,:)))))-
dot(r0(n,m+ns*nc,:),(r2(n,m+ns*nc,:)/norm(shiftdim(r2(n,m+ns*nc,:)))
));

Fac1a(1,1)=0;
Fac1a(1,2)=(panel(i,j).CP(3)-
panel(k,p).BV2(3))/((panel(i,j).CP(3)-panel(k,p).BV2(3))^2+(-
panel(k,p).BV(2)-panel(i,j).CP(2))^2);
Fac1a(1,3)=(-panel(k,p).BV2(2)-
panel(i,j).CP(2))/((panel(i,j).CP(3)-panel(k,p).BV2(3))^2+(-
panel(k,p).BV2(2)-panel(i,j).CP(2))^2);
Fac2a=-1+((panel(i,j).CP(1)-
panel(k,p).BV2(1))/norm(shiftdim(r1(n,m+ns*nc,:))));

Fac1b(1,1)=0;
Fac1b(1,2)=-(panel(i,j).CP(3)-
panel(k,p).BV1(3))/((panel(i,j).CP(3)-panel(k,p).BV1(3))^2+(-
panel(k,p).BV1(2)-panel(i,j).CP(2))^2);
Fac1b(1,3)=-(-panel(k,p).BV1(2)-
panel(i,j).CP(2))/((panel(i,j).CP(3)-panel(k,p).BV1(3))^2+(-
panel(k,p).BV1(2)-panel(i,j).CP(2))^2);
Fac2b=(-1+((panel(i,j).CP(1)-
panel(k,p).BV1(1))/norm(shiftdim(r2(n,m+ns*nc,:)))));

omega_bound(n,m+ns*nc,:)=Fac1*Fac2;
omega_a_inf(n,m+ns*nc,:)=Fac1a*Fac2a;
omega_b_inf(n,m+ns*nc,:)=Fac1b*Fac2b;

Fu(n,m+ns*nc)=omega_bound(n,m+ns*nc,1)+omega_a_inf(n,m+ns*nc,1)+omeg
a_b_inf(n,m+ns*nc,1);

Fv(n,m+ns*nc)=omega_bound(n,m+ns*nc,2)+omega_a_inf(n,m+ns*nc,2)+omeg
a_b_inf(n,m+ns*nc,2);

Fw(n,m+ns*nc)=omega_bound(n,m+ns*nc,3)+omega_a_inf(n,m+ns*nc,3)+omeg
a_b_inf(n,m+ns*nc,3);

%Code derived from NASA TN D-6142 (Vortex Lattice Fortran


186

function [] = FinalOutput(CL,CM, CDi, CD0, geo, airfoildata, output)


% calculate score and other outputs, post to the GUI

q = 1/2*geo.density*geo.V^2;

%Determine the stall angles of attack for root and tip airfoils
alpha_r = (geo.i_r + geo.alpha)*180/pi;
chord_r = geo.c_r;
thick_r = str2num(geo.root(end-1:end)); %Determine thickness of root
airfoil
Re_r = geo.V*geo.density*chord_r/geo.visc;
k = cell2mat(airfoildata{5}(geo.rootindex));
alpha_stall_r = k(1) + k(2)*log10(Re_r) + k(3)*log10(Re_r)^2;

alpha_t = (geo.i_r + geo.twist + geo.alpha)*180/pi;


chord_t = geo.c_r*geo.taper;
thick_t = str2num(geo.tip(end-1:end)); %#ok<ST2NM> %Determine thickness of
tip airfoil
Re_t = geo.V*geo.density*chord_t/geo.visc;
k = cell2mat(airfoildata{5}(geo.tipindex));
alpha_stall_t = k(1) + k(2)*log10(Re_t) + k(3)*log10(Re_t)^2;

eta = (2*(1:geo.ns)-1)/(2*geo.ns);
chord = chord_r + eta*(chord_t-chord_r);
alpha = alpha_r + eta*(alpha_t-alpha_r);
alpha_stall = alpha_stall_r + eta*(alpha_stall_t-alpha_stall_r);
thick = thick_r + eta*(thick_t-thick_r);

counter = 0; %Tracks the number of panels that are stalled


percentstalled = 0; %Tracks percentage of wing that is stalled
averagechord = 0; %Tracks average chord of stalled wing
averagealphapaststall = 0; %Tracks the average number of degrees past stall
averagethickness = 0; %Tracks the average thickness of the airfoil sections
past stall
averagestallalpha = 0; %Tracks the average stall angle
for i = 1:geo.ns
if alpha(i) > alpha_stall(i)
counter = counter + 1;
percentstalled = percentstalled + 1;
averagechord = averagechord + chord(i);
averagealphapaststall = averagealphapaststall + (alpha(i) -
alpha_stall(i));
averagethickness = averagethickness + thick(i);
averagestallalpha = averagestallalpha + alpha_stall(i);
end
end

if counter ~= 0 %Some part of the wing is stalled; prevents divide by zero


percentstalled = percentstalled/geo.ns;
averagechord = averagechord/counter;
averagealphapaststall = averagealphapaststall/counter*pi/180; %Convert
to radians because stall model was developed in radians
averagethickness = averagethickness/counter;
averagestallalpha = averagestallalpha/counter;
end
%Determine lift while accounting for lift after the stall angle of attack
c = averagethickness/12; %Thickness coefficient
187

d = 100*c; %Determined by trial and error to closely model the shape and
behavior of the lift curve
k = 400*c;
CL_stall = sum(CL.c+CL.alpha*averagestallalpha*pi/180); %Average max lift
coefficient
%The following equation is the result of modeling the lift
%coefficient of the airfoil after the stall angle of attack as a
%2nd order ODE with initial position of CL_stall and initial
%velocity of CL_alpha. See Stall Model.nb for Mathematica
%derivation
if percentstalled == 0
stallcorrection = 1;
else
stallcorrection = exp(0.5*(-d-sqrt(d^2-
4*k))*(averagealphapaststall))*(-2*CL.alpha-CL_stall*d+CL_stall*sqrt(d^2-
4*k))/(2*sqrt(d^2-4*k))...
+exp(0.5*(-d+sqrt(d^2-
4*k))*(averagealphapaststall))*(2*CL.alpha+CL_stall*d+CL_stall*sqrt(d^2-
4*k))/(2*sqrt(d^2-4*k));
end

% Effect Ground

h_c = geo.alti/geo.c_av;
omega =(1-(1.32*h_c))/(1.05+(7.4*h_c)); %equation NASA TN D970 page 7 &
NASA TM 77
e_g =1/(sqrt(1-omega));

% EFFECT ENDPLATE
h_b = geo.endplate;
o_e =(1.66*2*h_b)/(1+(1.66*2*h_b)); %equation from Paul E.Hemke Drag of
Wing with Endplate equation 20
e_p = 1/(sqrt(1-o_e));

%Lift calculation without ground effect & endplate

CL = ((1-
percentstalled)*(CL.c+CL.alpha*geo.alpha)+percentstalled*stallcorrection);
L = CL*q*geo.S;

%Lift calculation with ground effect & endplate

CLind = CL*e_g*e_p;
Lind = CLind*q*geo.S;

% Moment Calculation without ground effect & endplate


CM = (CM.c+CM.alpha*geo.alpha);
M = CM*q*geo.S;

% Moment Calculation with ground effect & endplate


CMind = CM*e_g*e_p;
Mind = CMind*q*geo.S;

%CD0 is profile drag of wing, S_Sref*Cfe is the zero-lift drag of the


%fuselage from section 2.9.1 of Anderson, Aircraft Performance and Design
188

%The sum of these two terms represents the total skin friction drag on the
%wing body or the zero-lift drag

%Drag without ground effect & endplate


cda = CDi+(CD0 + geo.S_Sref*geo.cf);
Dind = CDi*q*geo.S;
Dprofile = (CD0 + geo.S_Sref*geo.cf)*q*geo.S;

Dtotal = Dind + Dprofile; %Newton

%Drag with ground effect & endplate


d_g = 1-omega;
d_p = 1-o_e;

cdaind = cda*d_g*d_p;
Dindind = Dind*d_g*d_p;
Dprofileind = Dprofile*d_g*d_p;

Dtotalind = Dindind + Dprofileind; %Newton

%Set output
set(output.totaldrag,'String',num2str(round(Dtotal)));
set(output.inddrag,'String',num2str(round(Dind)));
set(output.profdrag,'String',num2str(round(Dprofile)));
set(output.lift,'String',num2str(L)); %Convert to lbf
set(output.CLaverage,'String',num2str(CL));
set(output.CDaverage,'String',num2str(cda));
set(output.Moment,'String',num2str(M));
set(output.CM,'String',num2str(-CM));
set(output.totaldrag2,'String',num2str(round(Dtotalind)));
set(output.inddrag2,'String',num2str(round(Dindind)));
set(output.profdrag2,'String',num2str(round(Dprofileind)));
set(output.lift2,'String',num2str(round(Lind))); %Convert to lbf
set(output.CLaverage2,'String',num2str(CLind));
set(output.CDaverage2,'String',num2str(cdaind));
set(output.Moment2,'String',num2str(Mind));
set(output.CM2,'String',num2str(-CMind));
189

APPENDIX E

ROUTH-HURWITZH STABILITY CRITERION CODE

%===================================================================
% Routh-Hurwitz Stability Criterion
%===================================================================

clc;
disp(' ')
D=input('Input coefficients of characteristic equation,i.e:[an an-1
an-2 ... a0]= ');
l=length (D);

disp(' ')
disp('----------------------------------------')
disp('Roots of characteristic equation is:')
roots(D)

%%=======================Program Begin==========================

% --------------------Begin of Bulding array------------------------


--------
if mod(l,2)==0
m=zeros(l,l/2);
[cols,rows]=size(m);
for i=1:rows
m(1,i)=D(1,(2*i)-1);
m(2,i)=D(1,(2*i));
end
else
m=zeros(l,(l+1)/2);
[cols,rows]=size(m);
for i=1:rows
m(1,i)=D(1,(2*i)-1);
end
for i=1:((l-1)/2)
m(2,i)=D(1,(2*i));
end
190

end

for j=3:cols

if m(j-1,1)==0
m(j-1,1)=0.001;
end

for i=1:rows-1
m(j,i)=(-1/m(j-1,1))*det([m(j-2,1) m(j-2,i+1);m(j-1,1) m(j-
1,i+1)]);
end
end

disp('--------The Routh-Hurwitz array is:--------'),m


% --------------------End of Bulding array--------------------------
------

% Checking for sign change


Temp=sign(m);a=0;
for j=1:cols
a=a+Temp(j,1);
end
if a==cols
disp(' ----> System is Stable <----')
else
disp(' ----> System is Unstable <----')
end

%=======================Program Ends==========================
191

APPENDIX F

SAVITSKY METHODE PROGRAM

Figure F.1 Planing Hull

INPUT

Hull LWL 0.914 metres 3.00 feet Length of waterline


B 0.129 metres 0.42 feet Beam
VCG 0.295 metres 0.97 feet VCG
6 kg 14.00 lbf Displacement
T 12.00 12.00 Deadrise @ transom
)0( 12.00 12.00 Deadrise @ amidships
L )0( 0.457 metres 1.500 feet Distance to amidships
0.00 0.00
0.00 0.00 Angle of thrust line
192

f 0.152 metres 0.50 feet Distance from thrust line to VCG

Minimum speed valid


Vmin 1.3 kn 2.2 feet/s
for this analysis
Maximum speed valid
Vmax 28.4 kn 48.0 feet/s
for this analysis

S/str LOA 1.165 metres 3.82 feet Length overall


Bmax 0.130 metres 0.42 feet Maximum beam
0.039 metres 0.13 feet Moulded depth of hull
HSS 0.147 metres 0.48 feet Height of house
BSS 0.123 metres 0.40 feet Breadth of house
2 2
ASS 0.018 m 0.20 feet Frontal area of house

Propulsion System

Number N 2 Number of propellers

Trim Tab cF 0.000 metres 0 feet Chord


0.000 ( <= 1 ) 0.000 Span ratio
0.0 0.0 Deflection angle

Rudder crudder 0.000 metres 0.00 feet Chord


t 0.000 metres 0.00 feet Thickness
2 2
Arudder 0.000 m 0.00 feet Area

xc 0.000 metres 0.00


feet Centrepoint
yc 0.000 metres 0.00 feet

Shaft shaft 0.000 metres 0.00 feet Diameter of shaft


l 0.000 metres 0.00 feet Length of shaft & hub

xc 0.000 metres 0.00


feet Centrepoint
yc 0.000 metres 0.00 feet

Strut cstrut 0.000 metres 0.00 feet Chord


t 0.000 metres 0.00 feet Thickness
2
Astrut 0.000 m 0.00 Area
feet2 ****
xc 0.000 metres 0.00 feet
Centrepoint
yc 0.000 metres 0.00 feet

Coding for iteration


Sub BerekenTrim()
Dim trim As Double
193

Dim i, j, kolom As Integer


Dim A(14), tijdelijk(14) As Double
Sheets("IO").Select
rij = 57
kolom = 4

j = -1
For i = 0 To 14 Step 1
Cells(rij + i, kolom).Value = 0.0001
If Worksheets("Berekenings").Cells(13, 3 + i).Value > 0 Then j = i
Next i

For i = (j + 1) To 14 Step 1
For trim = 0.1 To 15.1 Step 1
Cells(rij + i, kolom).Value = trim
If Worksheets("Berekenings").Cells(13, 3 + i).Value < 0 Then tijdelijk(i) =
Cells(rij + i, kolom).Value
If Worksheets("Berekenings").Cells(13, 3 + i).Value >= 0 Then Exit For
Next trim
Cells(rij + i, kolom).Value = tijdelijk(i)
A(i) = tijdelijk(i)
Next i

For i = (j + 1) To 14 Step 1
For trim = A(i) To 15.1 Step 0.1
Cells(rij + i, kolom).Value = trim
If Worksheets("Berekenings").Cells(13, 3 + i).Value < 0 Then tijdelijk(i) =
Cells(rij + i, kolom).Value
If Worksheets("Berekenings").Cells(13, 3 + i).Value >= 0 Then Exit For
Next trim
Cells(rij + i, kolom).Value = tijdelijk(i)
A(i) = tijdelijk(i)
Next i
194

For i = (j + 1) To 14 Step 1
For trim = A(i) To 15.01 Step 0.01
Cells(rij + i, kolom).Value = trim
If Worksheets("Berekenings").Cells(13, 3 + i).Value < 0 Then tijdelijk(i) =
Cells(rij + i, kolom).Value
If Worksheets("Berekenings").Cells(13, 3 + i).Value >= 0 Then Exit For
Next trim
Cells(rij + i, kolom).Value = tijdelijk(i)
A(i) = tijdelijk(i)
Next i

For i = (j + 1) To 14 Step 1
For trim = A(i) To 15.001 Step 0.001
Cells(rij + i, kolom).Value = trim
If Worksheets("Berekenings").Cells(13, 3 + i).Value < 0 Then tijdelijk(i) =
Cells(rij + i, kolom).Value
If Worksheets("Berekenings").Cells(13, 3 + i).Value >= 0 Then Exit For
Next trim
If Cells(rij + i, kolom).Value = " > 13 " Then Exit For
Cells(rij + i, kolom).Value = tijdelijk(i)
A(i) = tijdelijk(i)
Next i

For i = (j + 1) To 14 Step 1
For trim = A(i) To 15.0001 Step 0.0001
Cells(rij + i, kolom).Value = trim
If Worksheets("Berekenings").Cells(13, 3 + i).Value >= 0 Then Exit For
Next trim
If Cells(rij + i, kolom).Value = " > 13 " Then Exit For
Next i

If j <> -1 Then
For i = 0 To j Step 1
195

Cells(rij + i, kolom).Value = " - "


Next i
End If

For i = (j + 1) To 14 Step 1
If Cells(rij + i, kolom).Value > 15 Then Cells(rij + i, kolom).Value = " > 15 "
Next i

End Sub

OUTPUT

V LCG D T
[kn] [ft] [metres] [] [lbf] [kN] [lbf] [kN]
3 1.500 0.457 - - - - -
4 1.500 0.457 3.38 2 0.0 2 0.0
5 1.500 0.457 3.59 2 0.0 2 0.0
6 1.500 0.457 3.87 2 0.0 2 0.0
8 1.500 0.457 4.48 2 0.0 2 0.0
10 1.500 0.457 4.65 3 0.0 3 0.0
12 1.500 0.457 4.29 3 0.0 3 0.0
14 1.500 0.457 3.76 3 0.0 3 0.0
16 1.500 0.457 3.24 4 0.0 4 0.0
18 1.500 0.457 2.79 4 0.0 4 0.0
20 1.500 0.457 2.41 5 0.0 5 0.0
22 1.500 0.457 2.10 6 0.0 6 0.0
25 1.500 0.457 1.73 7 0.0 7 0.0
26 1.500 0.457 1.63 8 0.0 8 0.0
28 1.500 0.457 1.45 9 0.0 9 0.0
196

APPENDIX G

FREE RUNNING TEST SET UP

G.1 Instrument selection for free running test

The components and features are selected based on performance of WIG


craft during take-off involving thrust, motor RPM and control system. Motor
propeller has been selected to produce thrust during take-off, thrust meter to measure
the thrust force generated by the motor propeller, and the RPM sensor is selected to
compute the motor rotation per minute. Transmitter controller (remote control) is
used to control the movement of WIG craft model. Selection of telemetry receiver
also needs to receive signals from the transmitter controller and then sends a signal
to the servo motor to move the motor at the desired speed. The speed of the WIG
model during take off until it reaches and maintain in the air specifically controlled
by Electric Speed Control (ESC). Selection of the appropriate ESC is necessary in
order to properly process the signals and control the speed (timing) engine
RPM. Pitot tube is selected to a part of the instruments in this study because it
fulfills the criteria function to calculate the speed of the moving model. Based on the
performance of WIG craft model to be studied, such as speed, engine RPM, altitude,
and the take off distance, a data recorder has been selected for each of the readings
recorded during the take off and when it maintains in the air. Then, choosing the
second telemetry receiver has been made to receive all the data collected by the data
197

recorder so that it more easily to transfer to the laptop screen (PC). Telemetry
software should be installed first to the laptop before all the data can be read.
The selection of the instrument tools are due to what the expectation during
the flight test. For detail selection can be seen at Table G.1

Table G.1: Selection instrument For Free Running test

Component or Function Type and characteristic


instrument
Remote Control To control the moving of model Using Futaba 6EX
(Transmitter) by transfer the signal to the Transmitter with
telemetry receiver. the 2.4GHz FASST
system.
Very quickly
virtually
eliminating signal
conficts and
interruptions.
Easy Link system
features a unique
ID code in the
transmitter that
links to the
receiver.
198

Remote Control Receive the signal from the Using Futaba


(Receiver) remote control (transmitter). R607FS receiver.
Same frequency
with remote
control.
Powerful enough to
easily control giant-
scale models.
The R607FS uses
rubber grommets to
minimize stress on
the antennas.

Servo motor Ability to maintain a speed Servo Current


dictated by the velocity Monitor Micro
command input. Sensor by Eagle
Generate a proportional error tree.
signal equal to the difference The servo amplifier
between the velocity will respond
command and the actual proportionally to
motor speed. changes in the
Also acting as a source of velocity-input
electric power and given command (a -10V
voltan to brushless motor. to +10V signal).
Using the analog
velocity command
to drive the
amplifier at any
speed from reverse
to forward.
199

Electronic Speed ESC is a device that controls the ESCs are sized by
Control (ESC) motor RPM. Begin with throttle the amperage and
inputs are received by the radio voltage needed by
receiver and then sent to the your motor and
ESC to increase or decrease the propeller/rotor/fan
motor RPM. combination.
Micro planes the
ESCs are rated very
small at around 3.7
volts, 5 amps or
less whereas giant
scale ESCs go all
the way up to 50
volts and over 140
amps.

Motor Give the required thrust to the  Type of Brushless


model so that the model will motor producted by
move foward and fly. Maxx Product
 The removable
intake ring enables
the ducted fan to be
installed inside the
model or mounted
on the surface; it
improves the units
efficiency.
 Specification:
Input power: 1330
Watts.
Thrust: approx. 2.6 Kg,
5.7 lbs (92 oz).
200

RPM sensor Works with the Data Using Brushless


Recorder/eLogger to Motor RPM Sensor
measure RPM via pulses V2.
from any two of the wires The sensor will not
leading from Electronic work on Brushed
Speed Controller (ESC) to motor RPM.
motor. Simply connect a
Figure below show the RPM wire to the power
sensor. supply line.
For USB Data
Recorder and
MicroPower E-
Logger (starting
Version 2.0). RPM
sensor
range from 100 up
to approx. 50.000
RPM.
Thrust meter Measure the motor thrust Using Tahmazo
without installing the motor Thurst Meter.
onto model. Can measure the
Determine the right motor, thrust in different
propeller and even S.I. units (kg, lb,
combination of both before oz) and up to a
installing them onto model. maximum weight
of 9.9kg.
A motor mounting
plate is included for
motors from 12mm
to 50mm in
diameter.
Airspeed micro The MicroSensor is a Airspeed
sensor and pitot precision instrument that MicroSensor V3
tube uses a Prandtl style pitot- from Eagle Tree
201

static tube to measure product is


airspeed. choosing.
The Airspeed
Microsensor is
intended for use
exclusively in
model planes, boats
and cars.
It is extremely
unlikely that the
installation of the
MicroSensor will
affect the models
radio range or
control.

Data recorder May be separate to three Below shown the


main component that picture of types of data
have their own function: recorder produced by
Onboard Flight Data Eagle Tree.
Recorder - this unit
collects the data from
built-in and external
sensors.
Onboard Telemetry
Transmitter - this unit
sends the data from the
Recorder to the
Dashboard
Seagull Dashboard
Telemetry Receiver - this
unit displays telemetry
202

data, and sounds alarms


when problems occur.
Temperature To measure temperature  A device for
sensor of brushless motor and measuring sushu
humidity in the must have high
surrounding area when efficiency in terms
the flight test is of his senses.
conducted.  For USB Data
Temperature and Recorder and Micro
humidity affect the Power E-Logger.
performance of the  Temperature range
model during the flight up to 121C/250F.
test.This is due to air
pressure differences
result from differences in
moisture.
203

G.2 Calibration RIG and Calibration Servo

The calibration RIG has been done for measure the amount of thrust produced
by each DC motor. From this experiment will get data thrust for each RPM.
Calibration RIG is started by select base foundation, where in this experiment using
wood. This wood is used as a base to mount the motor and thrust meter. Thrust
meter and motor is also bound to the zinc before it is mounted on the surface of the
wood. The distance and radius of each hole was calculated in advance to ensure a
strong structure in addition to obtain the expected results of the experiment. The
distance between the holes in the wood represent the actual distance between the
thrust meter and the motor. The distance between the thrust meter and the motor in
this experiment is 10cm. This distance is suitable for wind power generated by the
motor will not be refracted too much. This can improve the accuracy of the thrust
meter readings. To ensure that the thrust meter can read the thrust force generated by
the motor as a whole, a zinc plate 1 along the 30cm and 21.5cm width has been
screwed in the pre-drilled holes thrust meter. The tahmazo thrust meter was mounted
on the zinc plate 1 with a size of 4mm diameter screw. The tahmazo thrust meter is
used because it is easy to mount and portable device that can measure the motor
thrust without installing the motor onto model.
ESC Pheonix ICE 75SB has been used to control some other component parts
according to specifications set. It has 6 series of wire that has different functions.
There are two wire + ve and -ve is connected to the pole + ve and -ve battery. This
process is actually done at the end after all the other devices already finished setup.
State of the battery must always be ON when you want to connect both the wire.
Meanwhile, a wire in the middle connected to a remote control (receiver). This wire
is plugged in the slot 3 from the upper. It also shows that ESC is able to receive
signals (data) from the remote control (transmitter) to increase the current flow rate
and the motor can rotate at higher speeds. Three wires (red, white, black) are
connected to the motor. The black wire on the ESC is connected to the red wire on
the motor. With this, the direction of motor rotation is clockwise, and this means that
the motor was like being in a state of breathing air from the forward direction and
remove them to back as the thrust. The thrust meter readings will show improvement
204

if the ESC is receiving signals from remote control (transmitter) to increase the
speed of motor rotation.
The Brushless Motor Sensor works with the Data Recorder/eLogger to
measure RPM via pulses from any two of the wires leading from Electronic Speed
Controller (ESC) to the motor. The sensor works with all known brushless motors.
There are two unconnected wires extending from the Sensor, labeled Wire 1 and
Wire 2. These wires need to be electrically connected to any 2 of the 3 wires
leading from ESC to the motor. The end of these wires can be soldered, clipped or
otherwise electrically connected to one of the motor wires. With most ESC/motor
combinations, the sensor should work correctly if only the Wire 1 wire is
connected, and the Wire 2 wire is left disconnected. To calibrate the RPM sensor,
its need to know the number of poles brushless motor has. The term poles refers
to the number of magnets in the motor (NOT the number of stator teeth, legs, or
hammerheads.) Brushless RPM sensor is connected to the connection between
ESC and motor.
First of all, before the ESC is connected to the motor, the RPM sensor wire is
inserted in the centre of the connection between ESC and motor. Wire 1 in the RPM
sensor wire is connected to the wire lead to motor (red) and ESC (black). Then it
was wrapped with plastic wrap to prevent current leakage while ensuring that the
RPM sensor can give the accurately reading with a maximum value of motor
rotation. For more detail Calibration RIG set up can be seen in Figure G.1- figure G
3.
205

Figure G.1: Calibration RIG set up

Figure G.2: Wiring set up for Calibration RIG


206

Figure G.3: Sensor set up for Calibration RIG

Calibration servo is done to measure the angle of rudder and the output of
servo. It is done when the entire devices are installing to WIG model. It involved
three main components (battery, servo, multi meter) to check the performance of
servo and rudder. The servo is move smoothly left and right and the multi meter will
show up the transfer function (output) value for corresponding angle of rudder. The
multi meter is stopped when the rudder is in maximum deflection. The protector is
placed in between the space of rudder to measure the maximum angle of deflection.
The servo output is recorded at this state. The rudder then to be set for 0, 10,
20, 30, 40 and maximum is 50. For each angle, it will give the different for
value of servo output. The angle of rudder will affect the radial turning for WIG
model when the WIG model moves at the certain speed. During the flight test, the
performance of servo and rudder is recorded. Detail configuration can be seen at
Figure G.4
207

Figure G.4: Calibration servo set up

G.3 Telemetry data record system set up

Telemetry data record system is a unit that sends the data from the Recorder to
the Dashboard. The Data Recorder is not powered when connecting the transmitter,
and the red wire of the transmitter cable is corresponds with the red dot on the port
label (to the right of the connector). The Seagull Transmitter cable is plugged into
the Expansion port on the transmitter is installed in the upper row of pins, as
shown in the figure. With the Telemetry Transmitter, the data obtained in this
experiment can be collected and sent online to the Dashboard. A transmitter antenna
position is determined in the vertical so that data can be transmitted easily without
any barriers and avoids errors.
208

Figure G.5 Telemetry data record set up

Figure G.6: Onboard installation


209

Figure G.6: Ready to Fly


210

APPENDIX H

SIMULATION PROGRAM

Figure H.1: Overview Simulation Program


211

Figure H.2: Input Output Schematic


212

Figure H.1: Trimaran WIG Dynamics Schematic