1 Perbandingan Metode Prediksi18303297 Main

IJNAOE242_proof ■ 22 February 2019 ■ 1/16
International Journal of Naval Architecture and Ocean Engineering xxx (xxxx) xxx
55
Contents lists available at ScienceDirect 56
57
International Journal of Naval Architecture and Ocean Engineering 58
59
journal homepage: http://www.journals.elsevier.com/ 60
international-journal-of-naval-architecture-and-ocean-engineering/ 61
62
63
64
65
1 Comparative study of prediction methods of power increase and 66
2 67
3 propulsive performances in regular head short waves of KVLCC2 using 68
4
5 CFD 69
70
6 71
7 Q6,1 Cheol-Min Lee a, Jin-Hyeok Seo a, Jin-Won Yu b, Jung-Eun Choi b, Inwon Lee a, * 72
8 a
Department of Naval Architecture and Ocean Engineering, Pusan National University, Busan, South Korea 73
9 b
Global Core Research Center for Ships and Offshore Plants, Pusan National University, Busan, South Korea 74
10 75
11 76
12 a r t i c l e i n f o a b s t r a c t 77
13 78
14 Article history: This paper employs computational tools to predict power increase (or speed loss) and propulsion per- 79
15 Received 6 November 2018 formances in waves of KVLCC2. Two-phase unsteady Reynolds averaged Navier-Stokes equations have
Received in revised form
80
16 been solved using finite volume method; and a realizable k-ε model has been applied for the turbulent 81
28 January 2019
17 closure. The free-surface is obtained by solving a VOF equation. Sliding mesh method is applied to
Accepted 19 February 2019 82
18 Available online xxx
simulate the flow around an operating propeller. Towing and self-propulsion computations in calm water
83
are carried out to obtain the towing force, propeller rotating speed, thrust and torque at the self-
19 84
propulsion point. Towing computations in waves are performed to obtain the added resistance. The
20 Keywords: 85
Power increase
regular short head waves of l/LPP ¼ 0.6 with 4 wave steepness of H/l ¼ 0.007, 0.017, 0.023 and 0.033 are
21 taken into account. Four methods to predict speed-power relationship in waves are discussed; Taylor 86
Speed loss
22 expansion, direct powering, load variation, resistance and thrust identity methods. In the load variation 87
Propulsion performances
23 Regular short head waves method, the revised ITTC-78 method based on the ‘thrust identity’ is utilized to predict propulsive 88
24 KVLCC2 performances in full scale. The propulsion performances in waves including propeller rotating speed, 89
25 CFD thrust, torque, thrust deduction and wake fraction, propeller advance coefficient, hull, propeller open 90
Taylor expansion
26 water, relative rotative and propulsive efficiencies, and delivered power are investigated. 91
Direct powering
27 © 2019 Production and hosting by Elsevier B.V. on behalf of Society of Naval Architects of Korea. This is an
Load variation 92
28 open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Resistance and thrust identity method 93
29 94
30 95
31 96
32 97
33 1. Introduction temperature, roughness (or fouling), rudder angle, drift angle, etc. 98
34 This involuntary speed loss is included by adding a sea margin. The 99
35 The ship speed in the actual seas decreases at constant power sea margin has a typical value between 15% and 20% on the 100
36 condition or the power increases to maintain ship speed constant. maximum continuous rotating power (Eide, 2015). 101
37 The prediction of power increase (or speed loss) of ocean-going A lot of researches have been performed to predict the RAW by 102
38 ships is essential to design hull form, to analyze speed sea trial computational or experimental methods. The RAW composes of 103
39 tests, to estimate the weather factor in EEDI formula (MEPC 66/ three components - vertical motions, diffraction and viscous effects 104
40 21Annex 5, 2014), fuel consumption and sea margin, and to find (Arribas, 2007). The first is known as drift force which has a second 105
41 the optimal seaway route for eco-operation. Speed loss in the actual order nature (Faltinsen, 1990), mainly caused by wave radiation via 106
42 seas depends on the resistance, propulsion, ship machinery, sea- the ship motion and becomes larger as the ship motions increase. 107
43 keeping, automatic control and maneuvering, etc. Speed losses are This is significant in the resonance frequency region of heave and 108
44 usually categorized as voluntary or involuntary (Faltinsen et al., pitch motions. The second is the diffraction force of the incident 109
45 1980). The former occurs when the ship master actively reduces waves on the ship which is primarily due to wave diffractions. This 110
46 the ship speed to avoid damage, danger or severe discomfort. The component is the most significant in short waves, where the ship 111
47 latter is due to the added resistance in seaway; wind, wave, water motion is nearly negligible. The third is damping forces associated 112
48 with the forced heave and pitch motions in calm water, and 113
49 insignificant compared to hydrodynamic damping of ship motions. 114
50 Q2 * Corresponding author. These three components involve energy dissipation. A major part of 115
51 E-mail address: inwon@pusan.ac.kr (I. Lee). this energy is transmitted to the wave radiating from the hull and a 116
Peer review under responsibility of Society of Naval Architects of Korea.
52 117
53 118
https://doi.org/10.1016/j.ijnaoe.2019.02.001
54 2092-6782/© 2019 Production and hosting by Elsevier B.V. on behalf of Society of Naval Architects of Korea. This is an open access article under the CC BY-NC-ND license 119
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Please cite this article as: Lee, C.-M et al., Comparative study of prediction methods of power increase and propulsive performances in regular
head short waves of KVLCC2 using CFD, International Journal of Naval Architecture and Ocean Engineering, https://doi.org/10.1016/
j.ijnaoe.2019.02.001
2 C.-M. Lee et al. / International Journal of Naval Architecture and Ocean Engineering xxx (xxxx) xxx
1 66
2 Nomenclature DPD power increase due to waves at constant speed 67
3 condition 68
4 DR external towing force 69
5 Alphabetical symbols Dt time step 70
6 B ship breadth DV speed loss due to waves at constant power condition 71
7 CAW added resistance coefficient in waves (¼ R2 AW2 ) z wave elevation 72
rgza B =LPP
8 CB block coefficient hD propulsive efficiency (¼hR ,hO ,hH ) 73
9 CF frictional resistance coefficient, non-dimensionlized hH hull efficiency [ ¼ ð1 tÞ=ð1 wÞ ] 74
10 by rg,S0 ,V 2 hO propeller-open-water efficiency 75
11 CN trial correction for propeller rate of revolution at hR relative rotative efficiency (¼Q0 =Q ) 76
12 speed l (a) wave length (b) scale ratio 77
13 CP trial correction for delivered power n fluid kinematic viscosity 78
14 CT total resistance coefficient in calm water, non- xi translatory and angular displacements in x, y, and z 79
15 dimensionlized by rg,S0 ,V 2 directions (i ¼ 1: surge, 2: sway, 3: heave: 4: roll, 5: 80
16 CV viscous resistance coefficient, non-dimensionlized by pitch, 6: yaw) 81
17 rg,S0 ,V 2 xn overload factor of propeller rotating speed 82
18 D (a) ship depth (b) diameter of propeller xP overload factor of propulsion efficiency 83
19 FD viscous force difference between model and ship r fluid density 84
20 scale pffiffiffiffiffiffiffiffiffiffiffiffiffi 4% percentage ratio of 4 in waves to 4 in calm water, 85
21 FN Froude number [ ¼ V= g,LPP ] where 4 ¼ DV, RT, n, T, Q, PD, t, w, hH, hH, hO, hR, hD, 86
22 fw weather factor in EEDI formula PD 87
23 g gravitational constant (¼9.81 m/sec2) u wave frequency 88
24 H wave height [m] V displacement volume 89
25 J propeller advance coefficient [ ¼ VA =ðn,DP Þ ] 90
26 k (a) wave number (¼2p/l) (b) form factor Subscripts 91
27 KT propeller thrust coefficient [ ¼ T=ðr,n2 ,DP 4 Þ ] a amplitude 92
28 KT% percentage ratio of the computed value of KTM to that e encounter wave 93
29 of the experiments M model scale 94
30 KQ propeller torque coefficient [ ¼ Q =ðr,n2 ,DP 5 Þ ] O (a) circular wave (b) normal self-propulsion 95
31 KQ% percentage ratio of the computed value of KQM to that condition (c) open water 96
32 of the experiments P propeller 97
33 kyy moment of inertia S ship scale 98
34 LCB longitudinal center of buoyancy forward of midship X external towing force for the load variation 99
35 LPP length between perpendiculars 100
36 LWL length at waterline Superscripts 101
37 n propeller rotating speed C calm water 102
38 p static pressure SP self-propulsion 103
39 PD delivered power W waves 104
40 PE effective power 105
41 Q propeller torque Acronyms 106
42 R (a) radius of propeller (b) resistance BN Beaufort number 107
43 RAW added resistance due to waves DOF degrees of freedom 108
44 RAWM% percentage ratio of the RAW to the RCTM EEDI energy efficiency design index 109
45 RN Reynolds number [ ¼ ðV,LPPÞ=n ] FS Fourier series 110
46 RT total resistance ITTC International Towing Tank Conference 111
47 R%T percentage ratio of the computational value to the JBC Japanese bulk carrier 112
48 theoretical KCS KRISO container ship 113
49 sm sea margin due to waves KRISO Korea Research Institute for Ship and Ocean 114
50 S0 wetted surface area Engineering 115
51 t thrust deduction fraction [ ¼ ðTM þ FD RTM C Þ=TM ] KVLCC2 KRISO very large crude oil carrier 2 116
52 T (a) draft (b) wave period (c) propeller thrust MEPC Marine Environment Protection Committee 117
53 TF towing force QNM torque and revolution method 118
54 U, W velocity component in (x,z) RANS Reynolds averaged Navier-Stokes 119
55 V ship speed RTIM resistance and thrust identity method 120
56 VA propeller advance speed SIMPLE semi-implicit method for pressure-linked equations 121
57 x,y,z coordinate SM sliding mesh method 122
58 w wake fraction TNM thrust and revolution method 123
59 VOF volume of fluids 124
60 Greek symbols 125
61 DCF roughness allowance coefficient 126
62 127
63 128
64 129
65 130
j.ijnaoe.2019.02.001
C.-M. Lee et al. / International Journal of Naval Architecture and Ocean Engineering xxx (xxxx) xxx 3
1 very small part of the energy is lost due to viscous friction. There- regular and irregular waves for a container ship, and discussed that 66
2 fore, RAW is considered a non-viscous phenomenon. This is the propeller open-water efficiency in irregular waves decreased 67
3 important when scaling from model experiments to full scale. with the increase of the wave height; the propeller rotating speed, 68
4 Gerritsma and Beukelman (1972) proposed CAW using quadratic thrust and torque considerably increased in regular head waves 69
5 wave amplitude; however, RAW does not show quadratic de- than those in regular following waves; the self-propulsion factors 70
6 pendency (Kim et al., 2017a). in waves are considered to be almost the same values as those in 71
7 The direct powering method in which the speed loss (or power calm water; the wake on the propeller plane is reduced in waves, 72
8 increase) in waves is directly predicted by reading off the ship especially close to the propeller hub and becomes more uniform 73
9 speed-delivered power curve, has commonly been used to analyze (Guo et al., 2012); the wake velocity (1-w) slightly increased as the 74
10 speed sea trial test results in ship yards because of the simplicity wave height increased; the self-propulsion factors in regular head 75
11 in spite of the lack of theoretical background. Journee (1976b) waves considerably varied with wave length, especially in the range 76
12 discussed that RAW was the largest part of the total resistance; of l/LPP ¼ 1.0e1.5 where ship motions were severe; the thrust 77
13 and predicted the speed loss at head sea conditions using empir- deduction fraction had a strong relationship with wave induced 78
14 ical formulas. Prpic-Orsic and Faltinsen (2009, 2012) discussed the motion; these added amounts had their peaks when the ship mo- 79
15 speed loss in actual sea states in the time domain by imple- tions were severe, showing the close resemblance with the RAW 80
16 menting a thrust loss model based on experimental data; and (Lee, 1983). Taskar et al. (2016) discussed that the wave direction Q3 81
17 developed a time domain numerical model to predict the speed had strong influence on the performance drop, with bow quar- 82
18 loss in irregular sea. Baree (2010) obtained the RAW using Maruo's tering and head sea conditions affecting propulsion performance 83
19 and Fujii-Takahashi's methods for Series 60; and discussed that the most. Valanto and Hong (2017) carried out the resistance and 84
20 the lower speeds/head and bow seas have more effect on the ship self-propulsion tests in seven wave directions with a cruise ship 85
21 performance than the higher speeds/other headings. Rathje et al. model; the highest RAW was measured in bow quartering seas not 86
22 (2011) studied a container ship in waves using RANS solver to in head seas; the propulsion losses due to ship motions appeared to 87
23 simulate the ship motion in head waves at the constant thrust become significant only with smaller ships at higher sea states; the 88
24 condition. Chuang and Steen (2011) explored the use of time- and average resistance in a sea area could be obtained by forming a 89
25 frequency-domain simulation in the prediction of speed loss in weighted sum over the resistance values in each sea state of the 90
26 waves; derived speed-loss formula using the Taylor expansion relevant wave scattergram. 91
27 method, and pointed out that the speed loss predicted from the Benchmark tests for the self-propulsion computations solving 92
28 model tests would be underestimated if towing force was not RANS equations of the KCS (3000 TEU container ship) and JBC (180k 93
29 taken into account; the speed loss reached peak value when wave DWT bulk carrier) models were carried out at the Tokyo 2015CFD 94
30 length approached ship length; and the speed loss increased lin- workshop (http://www.t2015.nmri.go.jp), in which the direct self- 95
31 early with increasing wave elevation in the long wave range propulsion computations were feasible to predict the resistance, 96
32 (Chuang and Steen, 2013). Three different methods to predict thrust and torque at the given propeller rotating speed and ship 97
33 power increase in irregular waves based on regular wave test re- speed. The self-propulsion point could be obtained applying load 98
34 sults are described in the ITTC-recommended procedures and variation method (Choi et al., 2009, 2010; Seo et al., 2010) or using a 99
35 guidelines 7.5-02-07-02.2 (2014); QNM, TNM, and RTIM. In the speed controller to modify the propeller rotating speed until the 100
36 RTIM, the resistance tests in waves, and the resistance, propeller- target speed was reached (Carrica et al., 2010; Castro et al., 2011). 101
37 open-water and self-propulsion tests in calm water are utilized Kayano et al. (2013) performed full-scale experiments of a 6,720 102
38 assuming thrust deduction and wake fractions in waves are the G.T. training ship, and clarified that the power increase in the waves 103
39 same as those in calm water. In the TNM, the self-propulsion tests was due to not only added resistance but also change of propulsive 104
40 in waves, and propeller-open-water and self-propulsion tests in efficiency; thrust measurement was important to calculate the 105
41 calm water are utilized assuming wake fraction in waves is the propulsive efficiency and the self-propulsion factors; wake coeffi- 106
42 same as that in calm water. In the QNM, only the self-propulsion cient was largely affected by ship's condition such as displacement 107
43 tests in waves are utilized; resistance and propeller-open-water and trim, not propeller loading. Taskar et al. (2016) suggested that 108
44 tests in calm water and in waves are unnecessary since propeller the sea margin include the effect of drop in propulsion efficiency in 109
45 rotating speed and torque variations are directly reflected to the waves. 110
46 delivered power. Kim et al. (2017b) employed the 2-D linear po- This paper discussed the speed loss (or power increase) and 111
47 tential theory to estimate the added resistance and ship speed loss propulsive performances in waves for KVLCC2 (300k DWT VLCC). 112
48 of the S175 container ship due to wind and waves; pointed out Towing and self-propulsion computations in calm water are applied 113
49 that the speed loss for the wave and wind directions from head to to obtain the resistance, propeller rotating speed, thrust and torque 114
50 bow seas is higher than the speed loss from beam directions and at the self-propulsion point. Towing computations in waves are 115
51 following sea directions. applied to obtain RAW. The regular short head waves of l/LPP ¼ 0.6 116
52 The propulsion point will change since the thrust has to balance with 4 wave steepness of H/l ¼ 0.007, 0.017, 0.023 and 0.033 are 117
53 the RAW. Journee (1976a) showed that the RAW quadratically varied taken into account. Computational methods and conditions are 118
54 with the wave amplitude; the added propeller rotating speed, described in Section 3. The results using four methods to predict 119
55 thrust and torque are linear to the RAW; the propulsive efficiency is speed-power relationships in waves are discussed in Section 4; 120
56 not influenced by the wave motion but only by a decreasing pro- Taylor expansion, direct powering, load variation, and RTIM 121
57 peller loading. When propeller is working close to the free surface, methods. In the load variation method, the revised ITTC-78 method 122
58 additional wave motion will be set up by the propeller and lead to based on the ‘thrust identity’ is utilized to predict propulsive per- 123
59 reduced thrust. Faltinsen et al. (1980) explained the ‘Wagner effect’, formances in full scale. The resistance and propulsion performances 124
60 which led to large thrust loss due to the ventilation and propeller in waves including propeller rotating speed, thrust, torque, thrust 125
61 out-of-water. As long as the propeller is fully submerged, the deduction and wake fraction, propeller advance ratio, hull, 126
62 ventilation and out-of-water effects can be excluded. Thus, the propeller-open-water, relative rotative and propulsive efficiencies, 127
63 propeller-open-water characteristics don't change in waves and delivered power are discussed. Viscous solver of STAR- 128
64 (McCarthy et al., 1961; Nakamura and Naito, 1975). Nakamura and CCM þ v11.04 is utilized. 129
65 Naito (1975) performed the resistance and self-propulsion tests in 130
j.ijnaoe.2019.02.001
1 2. Objective ship and calculation conditions 66

2 67
3 The objective ship is KVLCC2, which was designed by the KRISO 68
4 to provide data for both explication of flow physics and CFD vali- 69
5 dation for a modern 300k DWT tanker with bulbous bow and stern. 70
6 No full-scale ship exists. 71
7 The principal dimensions of ship (KVLCC2) at full-load draft and 72
Fig. 1. Body plans and side view of KVLCC2.
8 propeller (KP458) are listed in Table 1. 73
9 The design speed (VS) is 15.5 knots, which corresponds to 74
10 VM ¼ 1.136 m/s, FN ¼ 0.142 and RNM ¼ 7.113 106. 75
11 The body plans and side view of KVLCC2 are presented in Fig. 1. 76
12 The resistances in calm water and in waves are obtained from 77
13 the towing computations. The propulsive performances are ob- 78
14 tained from the towing and self-propulsion computations in calm 79
15 water. The regular short head waves of l/LPP ¼ 0.6 with 4 wave 80
16 steepness of H/l ¼ 0.007, 0.017, 0.023 and 0.033 have been taken 81
17 into account. The wave heights in full scale are 1.28, 3.16, 4.44 and 82
18 6.41 m with the same wave length of 192.0 m, which correspond to 83
19 BN 4, 6, 7 and 8, respectively. Note that the corresponding BF is for 84
20 the regular short waves, which may not be the same as that for the 85
21 short crested irregular waves. 86
22 At the towing computations, a rudder is not fitted to the model 87
23 ship; double-body model is used to obtain form factor; calm water and 88
24 regular short head wave conditions have been taken into account; the 89
25 heave and pitch motions are free; the range of the velocity is 14.0e15.5 90
Fig. 2. Coordinate system.
26 knots (VM ¼ 1.026e1.136 m/s) with 0.5 knots (DVM ¼ 0.037 m/s) in- 91
27 terval. At the self-propulsion computations, a propeller and a rudder 92
28 are fitted to the model ship; calm water condition is taken into ac- bow, sinkage to downward and trim by head are negative. 93
29 count; the heave and pitch motions are free; the range of the velocity 94
30 is 14.0e15.5 knots with 0.5 knots interval; 4 propeller rotating speeds 95
31 while keeping the ship speed constant are applied. 3.1. Theoretical background 96
32 97
33 The governing equations are the continuity and the two-phase 98
3. Computational methods and conditions unsteady RANS equations. A realizable k-ε model is applied for
34 99
35 the turbulent closure. The free-surface is obtained by solving a VOF 100
The ship is advancing at constant forward speed VS in sinusoidal equation. This study does not go into what the computational
36 101
waves with an arbitrary heading. This is assumed to be the same as method is, the details of which may be found in STAR-CCM þ user
37 102
a fixed ship with flow coming from upstream to down in sinusoidal guide (CD Adapco, 2016). The computations are carried out at the
38 103
waves with an encounter frequency. The ship fixed on a right- towing and self-propulsion conditions. The SM is employed to
39 104
handed coordinate system (x,y,z) is defined as positive x in the simulate the flow around an operating propeller. The time history
40 105
flow direction, positive y in the starboard, and positive z upward, as of the computational results for encounter wave elevation, resis-
41 106
shown in Fig. 2. The origin is the intersection point of the midship, tance, sinkage and trim are analyzed by FS.
42 107
the centerplane and the undisturbed free surface. Here, surge to To solve the governing equations, the flow domain is subdivided
43 108
44 into a finite number of cells and the equations are changed into 109
45 algebraic form via the discretization process. The cell-centered finite 110
Table 1
46 Principal dimensions of ship and propeller. volume method is used for the space discretization. The convective 111
47 terms are discretized using the second order upwind scheme. The 112
(a) Ship (KVLCC2)
48 diffusion terms utilize the central differences scheme. The SIMPLE 113
49
Ship Model algorithm is applied to the velocity-pressure coupling. The temporal 114
50 l 1.00 49.3 discretization is the second order Adams-Bashforth scheme. A high- 115
51 LPP [m] 320.0 6.491 resolution interface-capturing scheme to compute the evolution of 116
LWL [m] 325.5 6.602 free surface by solving VOF equation was employed.
52 B [m] 58.0 1.176 117
53 D [m] 30.0 0.609 118
54 T [m] 20.8 0.422 3.2. Computational domain and boundary conditions 119
55 ▽[m3] 312,622 2.609 120
S0 [m2] 27,194 11.189
56 CB 0.8098 0.8098
In the case of towing and self-propulsion computations, half and 121
57 LCB [%LPP] 3.48 3.48 full domains are used, respectively. The computational domain and 122
58 kyy/LPP 0.25 0.25 boundary conditions at the towing computations are present in 123
59 (b) Propeller (KP458) Fig. 3. The computational domain is a shape of rectangular 124
60 with 1.5 < x/LPP<2.5, 0 < y/LPP<1.5, and 1.0 < z/LPP<0.4. The 125
No. of propeller blade 4
61 DP [m] 0.200 fluid is water for the region of 1.0 < z/LPP<0.0 and air for 0.0 < z/ 126
62 Pitch-diameter ratio at 0.7 RP 0.721 LPP<0.5 at the initial condition. Grid damping zone is applied in the 127
63 Expanded area ratio 0.431 region of 2.5 < x/LPP<4.5 including numerical damping zone of 128
64 Hub-diameter ratio 0.155 3.8 < x/LPP<4.5 to ensure that no unwanted wave reflection occurs 129
Rotation direction clockwise
65 at the boundaries of the solution domain. 130
j.ijnaoe.2019.02.001
1 66
2 67
3 68
4 69
5 70
6 71
7 72
8 73
9 74
10 75
11 76
12 77
13 78
14 79
15 80
16 81
17 82
18 83
19 84
20 85
21 86
22 87
23 88
24 89
25 Fig. 3. Computational domain and boundary conditions at towing condition.
90
26 91
27 92
28 No-slip conditions are applied on the hull surface. Standard wall resolution. This is done using background grid, in which there is no 93
29 function is utilized. Velocity inlet and pressure outlet boundary hull. The under-relaxation factors are set to 0.2 for momentum and 94
30 conditions are applied for the inlet and outlet boundary plane, 0.8 for pressure. The time step is set 0.0025 s, corresponding to 1/ 95
31 respectively. Symmetry condition is applied on symmetry and side 500 of the Te, with the number of iterations per time step of 20. The 96
32 boundary plane. Slip wall boundary condition is applied on top and maximum physical time sets 60 s. 97
33 bottom boundary plane. The R%T for the wave length and height are listed in Table 2. The 98
34 The first order linear wave theory is applied for the inlet wave-generation test results are listed in Table 2. The wave lengths 99
35 boundary plane expressed as; are nearly same within differences of 0.9%. The wave heights show 100
36 differences within 5.0%. The computed wave amplitude is used for 101
37 zðx; tÞ ¼ za ,cosðk0 x ue tÞ (1) the non-dimensionalization. 102
38 Fig. 5 displays snapshot of the computational results and theo- 103
39 Uðx; z; tÞ ¼ u0 ,za ,ek0 z cosðk0 x ue tÞ (2) retical values expressed in Eq. (1). The time of t/Te ¼ 0.7 is randomly 104
40 selected. The waves are well damped in the grid and numerical 105
41 damping zones. 106
42 Wðx; z; tÞ ¼ u0 ,za ,ek0 z sinðk0 x ue tÞ (3) 107
43 108
44 u20 ,za ,ek0 z u20 ,z2a ,e2k0 z 3.5. Operating propeller test 109
45 pðx; z; tÞ ¼ cosðk0 x ue tÞ (4) 110
k0 2
46 The SM is applied to simulate the flow around an actuating 111
47 propeller. The computed results obtained from the moving refer- 112
48 3.3. Grid generation ence frame method are used as the initial condition. Before 113
49 applying the SM at the self-propulsion computations, the compu- 114
50 The overset grid system is applied, in which the overlapping tational parameters are selected through the case study at the 115
51 grids move with the ship motion, which is obtained using the 6- propeller-open-water computation at J ¼ 0.4, where 116
52 DOF model. In this work, only the 2-DOF is taken into account, VA¼VM¼1.136 m/s and nM ¼ 30.0 rps. Computational domain and 117
53 i.e., heave and pitch. view of grid generation are displayed in Fig. 6. The propeller block 118
54 The overset grid system is displayed in Fig. 4. The grids are of cylindrical shape, which dimension is 0.8DP in longitudinal- and 119
55 generated using STAR-CCM þ by means of the trimmed mesh 1.2DP in radial-direction, rotates with the same propeller rotating 120
56 technique. A prismatic layer is generated near the wall to resolve speed. The cell number is 0.68 M. 121
57 the boundary layer. Five layers are used. The distance of the 1st grid The KT and KQ are compared with those of the experiments 122
58 from the hull is yþz120. Fine grids are set near free-surface. More (KTM ¼ 0.1757 and 10KQM ¼ 0.2067 at J ¼ 0.4; SIMMAN 2008, http:// 123
59 than 50 cells per wave length and 20 cells per wave height for the l/ www.simman2008.dk/KVLCC/KVLCC2/kvlcc2_geometry.html). 124
60 L ¼ 0.6 and H/l ¼ 0.017 are used. Total grid number is 2.7 M. The results of the case studies using various time steps, sub- 125
61 iteration numbers and order of the Adams-Bashforth scheme are 126
62 3.4. Wave generation test listed in Table 3. The computations using larger Dt are utilized as 127
63 the initial condition for the computations using the smaller Dt, that 128
64 Wave generation tests are performed to know the mesh's is, the propeller rotation angle corresponding to the Dt changes to 129
65 quality and to find the computational parameters which give high 10.0 / 5.0/2.5 / 1.0 deg. 130
j.ijnaoe.2019.02.001
1 66
2 67
3 68
4 69
5 70
6 71
7 72
8 73
9 74
10 75
11 76
12 77
13 78
14 79
15 80
16 81
17 82
18 83
19 84
20 85
21 86
22 87
23 88
Fig. 4. Overset and trimmed grid system.
24 89
25 90
26 the experiments of the wave pattern, wave profile on the hull and 91
Table 2
27 Wave-generation test results. wave cuts at the towing computations in calm water. The compu- 92
28 tational results show good agreement with the experiments. 93
H/l l[m] H [m]
29 Therefore, the trimmed mesh on the free surface as shown in Fig. 4 94
30 Theory CFD R%T Theory CFD R%T 95
and cell numbers near free surface are applicable to simulating the
31 0.007 3.895 3.880 99.6 0.026 0.027 105.0 free surface around present model ship. 96
32 0.017 3.895 3.861 99.1 0.064 0.066 102.7
Fig. 8 displays the comparison of the computational results with 97
33 0.023 3.895 3.866 99.3 0.090 0.089 98.9 98
0.033 3.895 3.908 100.3 0.130 0.131 101.0 the experiments of the axial velocity contours and velocity vectors
34 on the propeller plane at the towing computations in calm water. 99
35 The dashed line denotes a circle of propeller plane. The shape of an 100
36 ‘island’ of low-speed region is well displayed not only in the 101
37 The minimum KT% is selected as a criterium since the magni- 102
computational results but also in the experiments. The nominal
38 tude of KT is much larger than that of KQ. The Dt is 0.00009 s cor- 103
wake of the computational result is 0.531, which is slightly lower
39 responding to the 1 propeller rotating angle, relaxation factor is 104
than that of the experiment of 0.561. Therefore, the present grid
40 0.2, and a sub-iteration number of 45 are employed. The propellers 105
system is applicable to simulating the flow around a propeller
41 rotate eight times. 106
under the self-propulsion condition.
42 At the self-propulsion computations, full domain is used. The 107
43 same grid system is used as that at the towing computation except 108
44 3.6. Verification and validation 109
adding propeller block as shown in Fig. 9. The cell number of the
45 propeller block is 1.0 M. 110
46 Verification of the computational results using the present grid 111
system is not performed. However, the computational results are The same values of relaxation factors, Dt and sub-iteration
47 numbers are used as those at the operating propeller test. The 112
48 validated by comparing with the experiments (Kim et al., 2001). 113
Fig. 7 displays the comparison of the computational results with propellers rotate four times. Fig. 10 displays the time histories of
49 114
50 115
51 116
52 117
53 118
54 119
55 120
56 121
57 122
58 123
59 124
60 125
61 126
62 127
63 128
64 129
65 Fig. 5. Snapshot of wave profile of wave-generation tests using CFD at t/Te ¼ 0.7 in l/L ¼ 0.6 and H/l ¼ 0.017. 130
j.ijnaoe.2019.02.001
1 66
2 67
3 68
4 69
5 70
6 71
7 72
8 73
9 74
10 75
11 76
12 77
13 78
14 79
15 80
16 81
17 82
18 83
Fig. 6. Computational domain and view of grid generation.
19 84
20 85
21 Table 3 86
22 Case study of the effect of time step, sub-iteration number and accuracy of the Adams-Bashforth scheme. 87
23 Case Rotation [degree] Dt [sec] Adams-Bashforth scheme KTM KT% 10KQM KQ% 88
24 89
0 10.0 0.00093 1st 0.1685 4.1 0.2071 0.2
25 1 5.0 0.00046 1st 0.1693 3.6 0.2081 0.7
90
26 2 2.5 0.00023 1st 0.1705 2.9 0.2096 1.4 91
27 3 1.0 0.00009 1st 0.1728 1.7 0.2126 2.9 92
28 4 5.0 0.00046 2nd 0.1700 3.2 0.2090 1.1 93
5 2.5 0.00023 2nd 0.1705 2.9 0.2096 1.4
29 94
6 1.0 0.00009 2nd 0.1741 0.9 0.2144 3.7
30 95
31 96
32 97
33 98
34 99
35 100
36 101
37 102
38 103
39 104
40 105
41 106
42 107
43 108
44 109
45 110
46 111
47 112
48 113
49 114
50 115
51 Fig. 7. Wave pattern around a ship, and wave profile on the hull and wave cuts.
116
52 117
53 118
54 propeller thrust and torque coefficients at the conditions of head waves are discussed. The self-propulsion computations in 119
55 nM ¼ 7.83 rps at the design speed. The time averaged values, which calm water are also discussed. Four methods to predict power in- 120
56 are the 0th terms in the FS of the resistance, thrust and torque are crease (or speed loss) are described and the results are compared; 121
57 used. Taylor expansion, direct powering, load variation, RTIM. 122
58 The calculation was done by 60 cores (2.67 GHz Xeon pro- 123
59 cessors) Linux Cluster. The computing times at the towing com- 4.1. Towing computations in calm water and in waves 124
60 putations in calm water and in waves, and self-propulsion 125
61 computations in calm water are 10 h, 24 h and 48 h, respectively. The RAW is the difference between the RTW and RCT at the same 126
62 ship speed. The resistance, sinkage and trim in calm water and in 127
63 4. Results regular head waves in model scale at l/LPP ¼ 0.6 with various 128
64 steepnesses at various ship speeds are listed in Table 4. Some of 129
65 The towing computations in calm water and in regular short results are compared with the experiments (Yu et al., 2017). Note 130
j.ijnaoe.2019.02.001
1 4.2. Self-propulsion computations in calm water 66

2 67
3 In order to predict speed performance, it is necessary to obtain a 68
4 self-propulsion point of a ship. The self-propulsion point may be 69
5 found in the results of the towing and the self-propulsion com- 70
6 putations under-and over-loaded propeller conditions for the given 71
7 VM (Choi et al., 2009, 2010; Seo et al., 2010). In model scale, TF, 72
8 which is the difference in force between the RSPTM and the TM, is zero 73
9 at the self-propulsion point. 74
10 75
11 TF ¼ RSP
TM TM (5) 76
12 77
13 In a full ship scale, however, the TF should be corrected by the FD 78
14 (Choi et al., 2009, 2010) as follows; 79
15 80
1
16 FD ¼ rS V 2 ½ð1 þ kÞðCFM CFS Þ DCF (6) 81
17 2 M M 82
18 Here, CF is obtained from the ITTC-1957 model-ship correlation 83
19 line (ITTC-recommended procedures and guidelines 7.5-02-03- 84
20 01.4). The form factor (1 þ k ¼ 1.230) is obtained from the towing 85
21 computation applying the double-body model; 86
22 87
23 1 þ k ¼ CVM =CFM (7) 88
24 Fig. 8. Axial velocity contours and velocity vectors on the propeller plane.
89
25 It is assumed that CVM is the same as the total resistance coef- 90
26 ficient of the double-body model. 91
The changes of the TF (DTF ¼ TFX-FD, where the FD denotes that
27 that the za is the computed value as discussed in Table 2. 92
28 for the normal self-propulsion condition) are selected as (ITTC- 93
The ship is sunken in calm water and in regular short head
recommended procedures and guidelines 7.5-04-01-01.2);
29 waves of l/L ¼ 0.6, where the magnitudes are nearly same along the 94
30 wave steepness. The ship is trimmed to bow in calm water and in 95
regular short head waves of l/L ¼ 0.6, where the magnitude in-
DRS TF FD
31 ¼ CX ¼ ½0:1 ; 0:0; þ 0:1; þ 0:2 (8) 96
32 creases as the wave steepness increases. The RAWM% for various RCTS RTM FD 97
33 model ship speeds in H/l ¼ 0.007 and 0.017 are 3.2e4.0% and 98
34 20.1e22.8%, respectively. As the wave steepness increases, the RAW 99
TFX ¼ FD ½0:1; 0:0; þ0:1; þ0:2,ðRTM C FDÞ (9)
35 rapidly increases (i.e., the RAWM% increases from 4.0% at H/ 100
36 l ¼ 0.007e72.6% at H/l ¼ 0.033). Note that the positive (X ¼ 3,4), negative (X ¼ 1) and zero (X ¼ 2 101
37 The CAW vs. wave steepness at the design speed in model scale is or FD) values in the square bracket denote over-, under-loaded and 102
38 displayed in Fig. 11. If the RAW has quadratic dependence on wave normal self-propulsion conditions, respectively. 103
39 amplitude as mentioned in the previous work (Gerritsma and The self-propulsion computations in calm water are carried out 104
40 Beukelman, 1972), the CAW becomes constant along the wave to obtain RSP
TM, TM and QM at four propeller rotating speeds including 105
41 steepness. However, the CAW does not show the quadratic de- a self-propulsion point while keeping the ship speed constant. Four 106
42 pendency. As the wave steepness increases, the quadratic de- model ship speeds of 1.026e1.136 m/s (14.0e15.5 knots in full scale) 107
43 pendency becomes weaker. The computational values of CAW are with 0.037 m/s interval (0.5 knots in full scale) are taken into 108
44 less than those of the experiments by 5.1e10.9% at the design account. 109
45 speed. Fig. 12 displays the procedure to obtain nM, TM and QM for 110
46 various TFs at the design speed. The values of the nM, TM and QM for 111
47 various TFs are obtained from reading off the self-propulsive 112
48 computational results. As the DRS increases, the TF decreases. The 113
49 nM, TM and QM increase due to the decreased TF to balance the over- 114
50 loaded propeller. 115
51 A speed-powering prediction in full scale is performed utilizing 116
52 the revised ITTC-78 method based on the ‘thrust identity’ (ITTC- 117
53 recommended procedures and guidelines 7.5-02-03-01.4; Choi 118
54 et al., 2009, 2010). The propeller-open-water characteristic curves 119
55 obtained from the experiments (http://www.simman2008.dk/ 120
56 KVLCC/KVLCC2/kvlcc2_geometry.html) are utilized. Resistance 121
57 and propulsive performances at normal load condition (DRS/ 122
58 RCTS ¼ 0.0) in full scale where CP ¼ 1.0 and CN ¼ 1.0 are listed in 123
59 Table 5. 124
60 The ratio of the hD at the over- and under-loaded conditions to 125
61 that at the normal load condition are plotted against the DR/RCT in 126
62 model and full scale in Fig. 13(a) and (b), respectively. The xP is the 127
63 slope of the linear curve going through {0,1} and fitted to the data 128
64 points with least square method. The effect on the propeller 129
65 Fig. 9. A view of grid generation near a propeller in self-propulsion computation. rotating speed (Dn/n0) against DPD/PDC is plotted in Fig. 13(c). The 130
j.ijnaoe.2019.02.001
1 66
2 67
3 68
4 69
5 70
6 71
7 72
8 73
9 74
10 75
11 76
12 77
13 78
14 79
15 80
16 81
Fig. 10. Time histories of propulsion coefficients at the conditions of nM ¼ 7.83 rps at the design speed.
17 82
18 83
19 Table 4
84
20 Total resistance, added resistance, sinkage and trim in calm water in regular head waves in model scale at l/LPP ¼ 0.6 with various wave steepnesses at various ship speeds. 85
21 86
VS [knots] VM [m/s] H/l zaM [m] RTM [N] RAWM [N] RAWM% CTMx103 or CAW Sinkage x103 [m] Trim [deg]
22 87
23 CFD Exp. (Yu et al., 2017) 88
24 14.0 1.026 0.000 0.000 24.32 e e 4.137 e 4.7 0.05 89
25 14.5 1.062 0.000 0.000 25.91 e e 4.109 e 5.2 0.05 90
15.0 1.099 0.000 0.000 27.63 e e 4.094 e 5.7 0.06
26 91
15.5 1.136 0.000 0.000 29.34 e e 4.072 e 6.3 0.07
27 92
28 14.0 1.026 0.007 0.013 25.09 0.77 3.2 2.241 e 0.6 0.03 93
14.5 1.062 0.007 0.013 26.80 0.89 3.4 2.609 e 0.7 0.03
29 15.0 1.099 0.007 0.013 28.66 1.03 3.7 2.955 e 0.8 0.02
94
30 15.5 1.136 0.007 0.013 30.53 1.19 4.0 3.436 3.864 0.5 0.02 95
31 96
14.0 1.026 0.017 0.032 29.87 5.55 22.8 2.843 e 1.9 0.07
32 14.5 1.062 0.017 0.032 31.56 5.65 21.8 2.886 e 1.6 0.07 97
33 15.0 1.099 0.017 0.032 33.40 5.77 20.9 2.921 e 1.7 0.06 98
34 15.5 1.136 0.017 0.032 35.24 5.90 20.1 3.050 3.242 1.6 0.06 99
35 15.5 1.136 0.023 0.045 39.42 10.08 34.4 2.834 3.002 2.1 0.09 100
36 15.5 1.136 0.033 0.065 50.63 21.29 72.6 2.601 2.737 3.6 0.14 101
37 102
38 103
39 4.3. Power increase or speed loss in waves 104
40 105
41 The DV is defined as the difference between the speed in calm 106
42 water and that in waves at constant power condition. And the DPD 107
43 is defined as the difference between the delivered power in waves 108
44 and that in calm water at constant speed condition: 109
45 110
46 DV ¼ V C V W at constant power (10) 111
47 112
48 113
49 DPD ¼ PDW PDC at constant speed (11) 114
50 The fw and sm are; 115
51 116
52 VW 117
53 fw ¼ at constant power (12) 118
VC
54 Fig. 11. Added resistance coefficient vs. wave steepness at the design speed in model 119
scale (l/LPP ¼ 0.6).
55 120
PDW PDC
56 sm ¼ 100½% at constant speed (13) 121
57 PDC 122
58 xn is the slope of the linear curve going through {0,0} and fitted to Four methods to predict speed-power relationship in waves are 123
59 the data points with least square method. The xP and xn are useful to discussed; Taylor expansion (Chung and Steen, 2011), direct pow- 124
60 correct propulsion efficiencies and shaft rotation rates of the sea ering, load variation, and RTIM methods. These methods are 125
61 trial tests as discussed in ITTC-recommended procedures and applicable to predict speed loss (or power increase) in waves when 126
62 guidelines 7.5-04-01-01.2. The xP and xn are also applicable to pre- towing and self-propulsion computations in calm water, and tow- 127
63 dict the delivered power and propeller rotating speed using Taylor ing computations in waves are available. Some of the results exceed 128
64 expansion and direct powering methods. the interpolation region. In this paper, the results within the 129
65 interpolation region are to be discussed. 130
j.ijnaoe.2019.02.001
1 66
2 67
3 68
4 69
5 70
6 71
7 72
8 73
9 74
10 75
11 76
12 77
Fig. 12. Procedure to obtain propeller rotating speed, thrust and torque for various propeller load conditions at the design speed.
13 78
14 79
15 Table 5 80
16 Resistance and propulsive performances at normal load condition in full scale. 81
17 82
VS [kts] PES [kW] PDS [kW] RTS [kN] nS [rpm] TS [kN] QS [kN-m] JS tS wS hHS hRS hOS hDS
18 83
19 14.0 10862 14679 1508 65.97 1858 2125 0.438 0.215 0.341 1.191 1.028 0.600 0.740 84
14.5 12010 16383 1610 68.37 1969 2288 0.443 0.206 0.333 1.191 1.016 0.605 0.733
20 15.0 13311 18309 1725 70.84 2099 2468 0.446 0.202 0.327 1.186 1.006 0.608 0.727
85
21 15.5 14645 20144 1837 73.28 2237 2625 0.447 0.201 0.324 1.183 1.009 0.609 0.727 86
22 87
23 88
24 89
25 90
26 91
27 92
28 93
29 94
30 95
31 96
32 97
33 98
34 99
35 100
36 101
37 102
38 103
39 104
40 105
41 106
42 Fig. 13. Overload factors of propulsion efficiency and propeller rotating speed. 107
43 108
44 109
45 4.4. Taylor expansion method 110
46 hW 111
RCTM þ RAW FDW ðRCTM FDC Þ hDM
RW
C
47 The can be expressed as a function of ship speed and
TM DVM ¼ V CM DM
(16) 112
approximated by Eq. (14): dRCTM
48 RCTM þ RAW FDW þ ð dV
þ dRdVAW ÞV CM 113
49 114
50 If we assumed that the towing force in calm water is the same as 115
51 that in waves (FDC¼FDW¼FD), Eq. (16) becomes 116
52 dRCTM dRAW 117
RW C
TM ¼ RTM þ RAW ð þ ÞDVM (14)
53 dV dV hW 118
RCTM þ RAW FD ðRCTM FDÞ hDM
C
54 The DVM is predicted at constant power condition in calm and in DVM ¼ V CM DM
(17) 119
dRCTM
55 waves (PDCM¼PDWM ) as expressed in Eq. (15): RCTM þ RAW FD þ ð dV
þ dRdVAW ÞV CM 120
56 121
57 The Eq. (17) is actually an expression for the speed loss due to 122
58 waves in model scale when the delivered power is kept constant. If 123
59 ðRCTM FDC Þ,V CM ðRW W W the overload factors (xp and xn) are available, the hW
D may be ob- 124
TM FD Þ,V M
60 ¼ (15) tained as expressed in Fig. 13(a); 125
hCDM hW
DM
61 126
62 Note that the TF should be applied to obtain correctly scaled hW R 127
63 propeller loading.
DM
¼ xPM AW þ1 (18) 128
hCDM RCTM
64 Inserting Eq. (14) into Eq. (15), and neglecting higher order 129
65 terms, then the DVM can be expressed as Eq. (16): The DVS can be obtained from the Froude's assumption. 130
j.ijnaoe.2019.02.001
1 4.6. Load variation method 66

pffiffiffi
2 DVS ¼ VS l,ðVM DVM Þ (19) 67
3 The RAW is assumed to be the external towing force, since the 68
4 The PDW (¼PDCþDPD) is obtained by reading off the speed- RAW is mainly caused by a non-viscous phenomenon. Then the FD 69
5 power curve at constant speed condition. And the nW
S is obtained in waves (FDW) can be expressed as; 70
6 as expressed in Fig. 13(c); 71
7 72
8 DPD FDW ¼ FD RAW (26) 73
DnW
S ¼ nS ðxn ,
C
Þ (20)
9 PDC 74
The self-propulsion point due to RAW may be found from the
10 If we assume that the propulsive efficiency in waves is the same towing and the self-propulsion computations for the given VM. The 75
11 as that in calm water (hWDM ¼ hDM), Eq. (17) becomes:
C
values of nM, TM, and QM at the self-propulsion point are obtained 76
12 from reading off those of the self-propulsive computational results 77
13 RAW at the value of TFW in the same way as discussed in Fig. 12. 78
14 DVM ¼ V CM (21) 79
dRCTM Fig. 14 displays towing force, propeller rotating speed, thrust
15 RCTM þ RAW FD þ ð dV
þ dRdVAW ÞV CM 80
and torque versus ship speed at self-propulsion point in model
16 The dR/dV can be obtained from the resistance curve, which is scale in calm water and in regular short head waves using load- 81
17 assumed to be the polynomial curve. Note that the speed loss in variation method. At model ship speed constant condition, as the 82
18 waves is to be underestimated if the FD is not taking into account as wave steepness increases, the TF decreases. The nM, TM and QM 83
19 shown in Eq. (21). increase due to the decreased TF to balance the over-loaded 84
20 The speed losses in regular head short wave of l/LPP ¼ 0.6 using propeller. 85
21 Taylor expansion method are listed in Table 6. A speed-powering prediction in full scale is performed utilizing 86
22 the revised ITTC-78 method as previously discussed. Fig. 15 displays 87
23 the resistance and propulsion characteristics versus ship speed in 88
24 calm water and in regular head waves (l/LPP ¼ 0.6) using load- 89
25 4.5. Direct powering method variation method. At constant ship speed condition, the resis- 90
26 tance in waves (R ¼ RCTSþDR) increases due to the RAW [Fig. 15(a)]; 91
27 The PDW
S is obtained as expressed in Eq. (22). the n, T and Q in waves increase to balance the propeller load 92
28 [Fig. 15(b), (c), (d)]. The t (¼1-R/T) decreases with increasing the R 93
29 94
RW W
TS ,V S (although T also increases, the increasing rate of R is greater than
30 PDW
S ¼ (22) 95
31
hW
DS
that of T) [Fig. 15(e)]. The w also decreases with similar tendency to
96
the t [Fig. 15(f)]. Thus, the hH [¼(1-t)/(1-w)] decreases with
32 decreasing the w (although the t also decreases, the decreasing rate 97
where
33 of w is greater than that of t) [Fig. 15(g)]. The J [ ¼ VS(1-w)/(n.DP)] 98
34 99
35 RW C
TS ¼ RTS þ RAWS (23) 100
36 Table 7 101
37 r Power and propeller-rotating-speed increases in regular head short waves of l/ 102
RAWS ¼ RAWM ,l , S 3
(24) LPP ¼ 0.6 at various ship speeds using direct powering method.
38 rM 103
39 H/l VS [knots] RAWS [kN] RAWS% DPD [kW] DPD% Dn [rpm] Dn% 104
40 0.007 14.0 95 6.3 1394 9.5 1.38 2.1 105
RAWS
41 hW
DS ¼ hDS ðxPS
C
þ 1Þ (25) 14.5 110 6.8 1656 10.1 1.53 2.2 106
42 RCTS 15.0 126 7.3 1952 10.7 1.72 2.4 107
15.5 146 7.9 2425 12.0 1.93 2.6
43 Note that the scale effect due to viscosity of the RAW is assumed 108
44 to be negligible. The Dn is obtained using Eq. (20)
0.017 14.0 683 45.3 12326 84.0 12.19 18.5 109
45 14.5 696 43.2 12573 76.7 11.60 17.0 110
The power increase in regular head short waves of l/LPP ¼ 0.6 at 15.0 710 41.2 12777 69.8 11.27 15.9
46 various ship speeds using direct powering method are listed in 111
15.5 726 39.5 14170 70.3 11.29 15.4
47 Table 7. The power increases by 9.5e12.0% for H/l ¼ 0.007 and 112
0.023 15.5 1241 67.6 28748 142.7 22.90 31.3
48 69.8e84.0% for H/l ¼ 0.017, where the resistance increases by 113
49 6.3e7.9% for H/l ¼ 0.007 and 39.5e45.3% for H/l ¼ 0.017.
0.033 15.5 2621 142.7 120935 600.4 96.35 131.5 114
50 115
51 116
52 117
Table 6
53 Speed loss in regular head short waves of l/LPP ¼ 0.6 at various ship speeds using Taylor expansion method.
118
54 119
H/l VS [knots] VM [m/s] RAWM [N] RAWM% DVM [m/s] DVM% DVS [kts] DVS%
55 120
56 0.007 14.0 1.026 0.77 3.2 0.020 2.0 0.27 1.9 121
57 14.5 1.062 0.89 3.4 0.022 2.0 0.30 2.1 122
15.0 1.099 1.03 3.7 0.024 2.2 0.33 2.2
58 123
15.5 1.136 1.19 4.0 0.028 2.4 0.37 2.4
59 124
0.017 14.0 1.026 5.55 22.8 0.135 13.2 1.84 13.1
60 125
14.5 1.062 5.65 21.8 0.128 12.1 1.75 12.1
61 15.0 1.099 5.77 20.9 0.125 11.4 1.71 11.4 126
62 15.5 1.136 5.90 20.1 0.130 11.4 1.76 11.4 127
63 0.023 15.5 1.136 10.08 34.4 0.210 18.5 2.86 18.5
128
64 129
0.033 15.5 1.136 21.29 72.6 0.390 34.3 5.31 34.3
65 130
j.ijnaoe.2019.02.001
1 66
2 67
3 68
4 69
5 70
6 71
7 72
8 73
9 74
10 75
11 76
12 77
13 78
14 79
15 Fig. 14. Towing force, propeller rotating speed, thrust and torque versus ship speed at self-propulsion point in model scale in calm water and in regular short head waves (l/ 80
LPP ¼ 0.6) using load-variation method.
16 81
17 82
18 83
19 84
20 85
21 86
22 87
23 88
24 89
25 90
26 91
27 92
28 93
29 94
30 95
31 96
32 97
33 98
34 99
35 100
36 101
37 102
38 103
39 104
40 105
41 106
42 107
43 108
44 109
45 110
46 111
47 112
48 113
49 114
50 115
51 116
52 117
53 118
54 119
55 120
56 121
57 122
58 123
Fig. 15. Resistance and propulsion characteristics versus ship speed in calm water and in regular head waves (l/LPP ¼ 0.6)using load-variation method.
59 124
60 125
61 decreases with increasing the n (although the w also decreases) hO and hR [Fig. 15(k)]. The delivered power rapidly increases to 126
62 [Fig. 15(h)]. Thus the hO decreases with the same tendency to the J maintain the ship speed constant as shown in Fig. 15(l) due to the 127
63 [Fig. 15(i)]. The hR (¼Q0/Q) decreases with increasing the Q RAW [Fig. 15(a)] and the decreased hD [Fig. 15(k)]. 128
64 [Fig. 15(j)]. The hD consequently decreases due to the reduced hH, To know these characteristics at constant delivered power 129
65 130
j.ijnaoe.2019.02.001
1 condition, it is necessary to investigate the resistance and propul- the n% moderately increases comparing to the RT%; the decreasing 66
2 sion characteristics versus delivered power (PD¼PDCþDPD) in calm rate of w% is greater than that t%, thus the hH% decreases; the hO% 67
3 water and in waves using load-variation method as displayed in decreases since the J% decreases due to the increasing n% (although 68
4 Fig. 16. The ship speed in waves rapidly decreases as the PD in- w% decreases); the hR% decreases since the Q% increases; thus, the 69
5 creases [Fig. 16(a)], the so called ‘speed loss in waves’. The RT, T, Q, t, hD% rapidly decreases since the hH%, hO% and hR% decrease; 70
6 w and hH are linear to the PD, that means the RT, T, Q, t, w and hH in totally, the PD% rapidly decreases due to not only increasing the RT 71
7 calm water are nearly same as those in waves at the constant PD % but also the decreasing the hD% (or not only increasing the n% but 72
8 condition [Fig. 16(b),(d),(e),(f),(g),(h)]. The n slightly decreases at also increasing Q%). Note that the DPD% is the wave component of 73
9 the constant PD condition as the RAW increases [Fig. 16(c)]. The J, hO, the sea margin, which mainly consists of wave, wind and fouling. 74
10 hR and hD significantly decreases at the constant PD condition as 75
11 the RAW increases [Fig. 16(i),(j),(k),(l)]. The J decreases as the RAW 76
12 increases due to the speed loss in waves (although the n decreases) 4.7. Resistance and thrust identity method (RTIM) 77
13 [Fig. 16(i)]. The hO shows similar tendency to the J [Fig. 16(j)]. The hR 78
14 decreases as the RAW increases [Fig. 16(k) since the Q0 decreases The power increase in waves is calculated according to the ITTC- 79
15 due to the decreased J. The hD consequently decreases due to the recommended procedures and guidelines 7.5-02-07-02.2 (2014). 80
16 reduced hO and hR [Fig. 16(l)]. Fig. 18 displays the procedure to obtain PDW S at constant ship 81
17 The effect of wave steepness on the propulsion performances in speed condition utilizing the propeller-open-water characteristic 82
18 regular head waves (l/LPP ¼ 0.6) at design speed is present in curves applying the RTIM. 83
19 Fig. 17. As the wave steepness increases, the RT% rapidly increases as The RW TS is obtained from Eq. (23). The t and w in waves are 84
20 discussed in Table 4 and Fig. 11; the T% and Q% are linear to the RT%; assumed to be the same as those in calm water, that is, tW C
S ¼ tS ¼ tS 85
21 and wW C 2
S ¼ wS ¼ wS. Then thrust in waves and KT/J can be obtained; 86
22 87
23 88
24 89
25 90
26 91
27 92
28 93
29 94
30 95
31 96
32 97
33 98
34 99
35 100
36 101
37 102
38 103
39 104
40 105
41 106
42 107
43 108
44 109
45 110
46 111
47 112
48 113
49 114
50 115
51 116
52 117
53 118
54 119
55 120
56 121
57 122
58 123
59 124
60 125
61 126
62 127
63 128
64 Fig. 16. Resistance and propulsion characteristics versus delivered power in calm water and in regular head waves (l/LPP ¼ 0.6)using load-variation method. 129
65 130
j.ijnaoe.2019.02.001
1 66
2 67
3 68
4 69
5 70
6 71
7 72
8 73
9 74
10 75
11 76
12 77
13 78
14 79
15 80
16 81
17 82
Fig. 17. Effect of wave steepness on the resistance and propulsion performances in regular head waves (l/LPP ¼ 0.6) at design speed.
18 83
19 84
20 coefficient KP is obtained on the KP curve; see Fig. 14(C). By using 85
21 this KP value, the power in waves is calculated by 86
22 87
23 KQ 88
24 KP ¼ (30) 89
25 J 3S 90
26 91
27 3 92
S ¼ 2p,nS ,QS ¼ 2p,KP ,rS ,ð1 wS Þ ,VS ,DP
PDW 3 2
(31)
28 93
29 The power increase in regular head short waves of l/LPP ¼ 0.6 at 94
30 various ship speeds using the RTIM are listed in Table 8. 95
31 96
32 97
33 98
34 4.8. Comparison of the results using four methods 99
Fig. 18. Propeller-open-water characteristic curves applying RTIM.
35 100
36 The delivered powers and propeller rotating speeds for various 101
37 ship speeds in calm water and in regular head waves using four 102
38 RW methods are compared in Fig. 19. At the condition of 103
39 TW
S ¼
TS
(27) PD ¼ 20,144 kW in regular head waves of l ¼ 192.0 m and 104
1 tS
40 H ¼ 1.28 m corresponding to BN ¼ 4 [Fig. 19(a)], the ship speed is 105
41 . 15.50 knots (design speed) with propeller rotating speed of 106
TW
42 KT J 2 ¼ S
2
(28) 73.28 rpm in calm water; 15.13 knots using Taylor expansion 107
43 rS ,D2P ,V 2S ,ð1 wS Þ method, 14.97 knots with 72.40 rpm using direct powering method, 108
44 14.96 knots with 72.44 rpm using load variation method, and 15.06 109
45 On the KT/J2 curve, JS is obtained; see Fig. 18(A) and (B). 110
knots with 72.54 rpm using RTIM. In regular head waves of
46
ð1 wS Þ,VS l ¼ 192.0 m and H ¼ 3.16 m corresponding to BN ¼ 6 [Fig. 19(b)], the 111
47 JS ¼ (29) delivered power to maintain the ship speed constant needs 1.70 112
nS ,DP
48 times using direct power method (34,314 kW), 1.62 times using 113
49 The nS is calculated by using Eq. (29). At this JS value, power load variation method (32,724 kW) or 1.50 times using RTIM 114
50 (30,257 kW) of that in calm water (20,114 kW). Note that the 115
51 increased delivered power in l ¼ 192.0 m and H ¼ 3.16 m is due to 116
Table 8
52 not only the increased resistance (39.5% of the RCTS) but also the 117
Power and propeller-rotating-speed increases in regular head short waves of l/
53 LPP ¼ 0.6 at various ship speeds using RTIM method. reduced propulsive efficiency (14.2% of the hCD). 118
54 The magnitude order of the PD in waves of l/LPP ¼ 0.6 and H/ 119
H/l VS [knots] RAWS [kN] RAWS% DPD [kW] DPD% Dn [rpm] Dn%
55 l ¼ 0.017 is Taylor expansion < RTIM < load variation < direct 120
56 0.007 14.0 95 6.3 1425 9.7 1.41 2.1 powering method. In waves of l/LPP ¼ 0.6 and H/l ¼ 0.007, the 121
57 14.5 110 6.8 1418 8.7 1.31 1.9 differences between the load variation and the direct powering 122
15.0 126 7.3 1578 8.6 1.39 2.0
58 method are very small. This is seemed to be due to the assumption 123
15.5 146 7.9 1884 9.4 1.50 2.0
59 that the RAW acts on the external towing force in load variation 124
0.017 14.0 683 45.3 9607 65.4 9.50 14.4
60 method. 125
14.5 696 43.2 9602 58.6 8.86 13.0
61 15.0 710 41.2 9751 53.3 8.60 12.1 The fw and sm in regular head waves of l/LPP ¼ 0.6 and H/ 126
62 15.5 726 39.5 10113 50.2 8.06 11.0 l ¼ 0.007 corresponding to BF ¼ 4 using four methods are sum- 127
63 0.023 15.5 1241 67.6 19425 96.4 19.12 26.1
marized in Table 9. In the case of l/LPP ¼ 0.6 and H/l ¼ 0.017 cor- 128
64 responding to BF ¼ 6, the fws are out of range at PD ¼ 20,144 kW as 129
0.033 15.5 2621 142.7 36768 182.5 26.37 36.0
65 shown in Fig. 19 (b). 130
j.ijnaoe.2019.02.001
1 66
2 67
3 68
4 69
5 70
6 71
7 72
8 73
9 74
10 75
11 76
12 77
13 78
14 79
15 80
16 81
17 82
18 83
19 84
20 85
21 86
22 87
23 88
24 89
25 90
26 91
27 92
28 93
29 94
30 95
Fig. 19. Ship speed-power-propeller rotating speeds in calm water and in regular head waves using four methods (l/LPP ¼ 0.6).
31 96
32 97
33 Table 9 hull efficiency are scarcely varied; propeller advance coefficient, 98
34 Whether factor and sea margin in regular head waves using four methods: l/ propeller-open-water, relative rotative and propulsive effi- 99
35 LPP ¼ 0.6 and H/l ¼ 0.007. 100
ciencies, and ship speed significantly decrease.
36 VS [knots] Taylor Direct Load variation RTIM ∙ As the wave steepness increases; quadratic dependency of 101
37 expansion powering added resistance due to waves becomes weaker; delivered po- 102
38 fw sm [%] fw sm [%] fw sm [%] fw sm [%] wer more rapidly decreases due to not only increasing added 103
39 resistance but also decreasing the propulsive efficiency. 104
14.0 0.981 6.7 e 9.5 e 10.4 e 9.7
40 14.5 0.979 7.5 0.971 10.1 0.969 10.4 0.971 8.7
105
41 15.0 0.978 7.2 0.971 10.7 0.970 11.0 0.975 8.6 106
42 15.5 0.976 e 0.966 12.0 0.965 11.9 0.972 9.4 107
Uncited references Q5
43 108
44 109
ITTC, 2008; ITTC, 2014; ITTC, 2018.
45 5. Conclusions 110
46 111
Acknowledgements
47 ∙ Power increase (or speed loss) and propulsion performances in 112
48 regular short head waves with varying levels in steepness of 113
49 This work was supported by the National Research Foundation Q4 114
KVLCC2 have been predicted through computational tools.
of Korea grant funded by the Korea government (MSIT) (No. 2011-
50 ∙ A comparative study utilizing four prediction methods has been 115
51 0030013) and by the Industrial Strategic Technology Development 116
performed -Taylor expansion, direct powering, load variation,
52 Program (No. 10076439) funded by the Ministry of Trade, Industry 117
resistance and thrust identity methods. The magnitude order of
and Energy, to which deep gratitude is expressed.
53 the increased delivered power in waves of l/LPP ¼ 0.6 and H/ 118
54 l ¼ 0.017 is Taylor expansion < RTIM < load variation < direct 119
55 powering method. References 120
56 ∙ At constant ship speed condition in waves; resistance, propeller 121
Arribas, F.P., 2007. Some methods to obtain the added resistance of a ship advancing
57 rotating speed, thrust and torque increase to balance over- 122
in waves. Ocean Eng. 34, 946e955.
58 loaded propellers. As a result, thrust deduction and wake frac- Baree, M.S., 2010. An investigation of ship performance in seas. In: Proceedings of 123
59 tion, hull efficiency, propeller advance coefficient, propeller- the International Conference on Marine Technology (MARTEC), 11-12, 124
60 December 2010, BUET, Dhaka, Bangladesh. 125
open-water, relative rotative and propulsive efficiencies Carrica, P.M., Castro, A.M., Stern, F., 2010. Self-propulsion computations using a
61 decrease; and the delivered power significantly increases. speed controller and a discretized propeller with dynamic overset grids. J. Mar. 126
62 ∙ At constant power condition in waves; resistance, thrust and Sci. Technol. 15, 316e330. 127
63 Castro, A.M., Carrica, P.M., Stern, F., 2011. Full scale self-propulsion computations 128
torque are scarcely varied; propeller rotating speed slightly
using discretized propeller for the KRISO container ship KCS. Comput. Fluids 51,
64 decreases. As a result, thrust deduction and wake fraction, and 35e47. 129
65 CD Adapco, 2016. STAR-CCMþ User Guide, version 11.04. 130
j.ijnaoe.2019.02.001
1 Choi, J.E., Kim, J.H., Lee, H.G., Choi, B.J., Lee, D.H., 2009. Computational predictions of International Journal on Marine Navigation and Safety of Sea Transportation 7 29
ship-speed performance. J. Mar. Sci. Technol. 14 (3), 322e333. (4), 521e526.
2 30
Choi, J.E., Min, K.-S., Kim, J.H., Lee, S.B., Seo, H.W., 2010. Resistance and propulsion Kim, M., Hizirb, O., Turana, O., Incecika, A., 2017a. Numerical studies on added
3 characteristics of various commercial ships based on CFD results. Ocean Eng. 37, resistance and motions of KVLCC2 in head seas for various ship speeds. Ocean 31
4 549e566. Eng. 140, 466e476. 32
5 Chuang, Z., Steen, S., 2011. Prediction of speed loss of a ship in waves. In: Pro- Kim, M., Hizirb, O., Turana, O., Incecika, A., 2017b. Estimation of added resistance 33
ceedings of the 2nd International Symposium on Marine Propulsors (SMP), M. and ship speed loss in a seaway. Ocean Eng. 141, 465e476.
6 Abdel-Maksoud. Hamburg, Germany, 1: 8. Kim, W.J., Van, D.H., Kim, D.H., 2001. Measurement of flows around modern com- 34
7 Chuang, Z., Steen, S., 2013. Speed loss of a vessel sailing in oblique waves. Ocean mercial ship models. Exp. Fluid 31, 567e578. 35
8 Eng. 64, 88e99. McCarthy, J.H., Norley, W.H., Ober, G.L., 1961. The Performance of a Submerged 36
Eide, E., 2015. Calculation of Service and Sea Margins. Master thesis. Institute for Propeller in Regular Waves. David Taylor Model Basin, Washington DC, USA.
9 Marine Technology Norwegian University of Science and Technology. MEPC 66/21 Annex 5, 2014. Guidelines on the Method of Calculation of the Attained
37
10 Faltinsen, O.M., 1990. Sea Loads on Ships and Offshore Structures. Cambridge Uni- Energy Efficiency Design Index (EEDI) for New Ships. 38
11 versity Press, Cambridge. Nakamura, S., Naito, S., 1975. Propulsive performance of a container ship in waves. 39
Faltinsen, O.M., Minsaas, K.J., Liapis, N., Skjørdal, S.O., 1980. Prediction of resistance J. Kansai Soc. Nav. Archit. Jpn. 158, 24e47.
12 and propulsion of a ship in a seaway. In: Proceedings of the 13th Symposium on Prpic-Orsic, J., Faltinsen, O.M., 2009. Speed loss calculation in a seaway. In: Pro-
40
13 Naval Hydrodynamics (SNH), pp. 505e529. ceedings of the 13th Congress. International Maritime Association of the Med- 41
14 Gerritsma, J., Beukelman, W., 1972. Analysis of the resistance increase in waves of a iterranean (IMAM), p. 393-400. 42
fast cargo ship. Int. Shipbuild. Prog. 19, 285e293. Prpic-Orsic, J., Faltinsen, O.M., 2012. Estimation of ship speed loss and associated
15 43
Guo, B., Steen, S., Deng, G.B., 2012. Seakeeping prediction of KVLCC2 in head waves CO2 emissions in a seaway. Ocean Eng. 44, 1e10.
16 with RANS. Appl. Ocean Res. 35, 56e67. Rathje, H., Schellin, T.E., Brehm, A., 2011. Speed loss in waves and wave-induced 44
17 ITTC, 2008. ITTC-recommended Procedures and Guidelines. Performance, Propul- torsion of a wide-breadth containership. Journal of Engineering for the Mari- 45
sion 1978 ITTC Performance Prediction Method 7.5-02-03-01.4. time Environment 225, 387e401.
18 46
ITTC, 2014. ITTC-recommended Procedures and Guidelines. Performance, Prediction Seo, J.H., Seol, D.M., Lee, J.H., Rhee, S.H., 2010. Flexible CFD meshing strategy for
19 of Power Increase in Irregular Waves from Model Test 7.5-02-07-02.2. prediction of ship resistance and propulsion performance. International Journal 47
20 ITTC, 2018. ITTC-recommended Procedures and Guidelines. Analysis of Speed/po- of Naval Architecture and Ocean Engineering 2, 139e145. 48
21 wer Trial Data ITTC 7.5-04-01-01.2. Taskar, B., Yum, K.K., Steen, S., Pedersen, E., 2016. The effect of waves on engine- 49
Journee, J.M.J., 1976a. Report 428, May 1976. Motions, Resistance and Propulsion of propeller dynamics and propulsion performance of ships. Ocean Eng. 122,
22 a Ship in Regular Head Waves, vol. 2. Delft University of Technology, Ship Hy- 262e277. 50
23 dromechanics Laboratory, Mekelweg, CD Delft, The Netherlands, p. 2628. Valanto, P., Hong, Y., 2017. Wave added resistance and propulsive performance of a 51
24 Journee, J.M.J., 1976b. Prediction of speed and behaviour of a ship in a seaway. Int. cruise ship in waves. In: Proceedings of the 27th International Ocean and Polar 52
Shipbuild. Prog. 23 (265), 1e24. Engineering Conference (ISOPE), 25-30 June, San Francisco, California, USA.
25 Kayano, J., Yabuki, H., Sasaki, N., Hiwatashi, R., 2013. A study on the propulsion Yu, J.W., Lee, C.M., Choi, J.E., Lee, I.W., 2017. Effect of ship motions on added resis-
53
26 performance in the actual sea by means of full-scale experiments. The tance in regular head waves of KVLCC2. Ocean Eng. 146, 375e387. 54
27 55
28 56
j.ijnaoe.2019.02.001

1 Perbandingan Metode Prediksi18303297 Main

Hochgeladen von

Dokumentinformationen

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

1 Perbandingan Metode Prediksi18303297 Main

Hochgeladen von

Copyright:

Verfügbare Formate

IJNAOE242_proof ■ 22 February 2019 ■ 1/16

1 2. Objective ship and calculation conditions 66

1 4.2. Self-propulsion computations in calm water 66

1 4.6. Load variation method 66

Das könnte Ihnen auch gefallen