표제지
초록
목차
1. 서론 26
1.1. 연구배경 및 내용 26
1.2. 논문구성 34
2. 잠수함 스텔스 성능 분석기법 36
2.1. 잠수함 스텔스 함형 설계 추세 36
2.2. 잠수함 음향표적강도 분석기법 50
2.2.1. 음향표적강도 정의 50
2.2.2. 잠수함 음향표적강도 해석기법 58
2.3. 잠수함 수중방사소음 분석기법 65
2.3.1. 잠수함 유동장 해석기법 67
2.3.2. 음향상사기법 70
2.3.3. 주파수 기반 음향상사를 이용한 유동소음 해석기법 74
2.4. 심층강화학습을 이용한 최적화 기법 81
2.4.1. 가치기반방법 85
2.4.2. 정책기반방법 89
3. 음향표적강도 저감을 위한 함형 설계기법 93
3.1. 음향표적강도 저감형 함형 설계기법 개발 95
3.1.1. 대상 함형 및 함형 변형 모델 97
3.1.2. 잠수함 음향표적강도 분석 모델 103
3.1.3. 잠수함 조종성능 분석 모델 107
3.1.4. 잠수함 함형 기하학적 제한조건 120
3.2. 심층강화학습을 이용한 음향표적강도 저감형 함형 최적화 123
3.2.1. 심층강화학습을 이용한 함형 최적화 기법 개발 123
3.2.2. 음향표적강도 저감형 최적 함형 도출 141
4. 수중방사소음 저감을 위한 와류저감장치 설계기법 159
4.1. 잠수함 함형 주위 유동특성 161
4.1.1. 부가물 주위 생성 와류특성 161
4.1.2. 부가물 와류에 의한 추진기 소음특성 변화 169
4.1.3. 부가물 주위 와류 자체소음 특성 172
4.1.4. 잠수함 주위 생성 와류특성 176
4.2. 선체-부가물-추진기 상호작용 소음해석 모델 개발 183
4.2.1. 수중환경 주파수 기반 음향상사 모델 183
4.2.2. 수중 추진기 소음해석 모델 검증 192
4.3. 와류저감장치 유동소음 저감 효과 분석 198
4.3.1. 잠수함 선체-부가물-추진기 상호작용 유동소음해석 198
4.3.2. 와류저감장치 장착에 따른 유동장 변화 분석 223
4.3.3. 와류저감장치 장착에 따른 유동소음 변화 분석 236
4.4. 와류저감장치 형상 최적화 250
4.4.1. 부가물 와류 상사특성 분석 251
4.4.2. 와류저감장치 최적형상 설계기준 도출 258
5. 잠수함 스텔스 함형설계시스템 개발 283
5.1. 심층강화학습을 이용한 스텔스 함형 최적화 시스템 285
5.2. 심층강화학습을 이용한 스텔스 검토 함형 도출 287
5.3. 스텔스 성능 종합 검토 및 최적 함형 도출 300
6. 결론 및 향후 추천 연구 327
6.1. 결론 327
6.2. 향후 추천 연구 329
부록 A. VRD 설계기준 도출 332
A.1. 전면부 VRD 설계기준 도출 332
A.2. 후면부 VRD 설계기준 도출 337
참고 문헌 341
Abstract 350
Table 1. Comparisons of shape optimization studies using DRL 94
Table 2. Comparison of hydrodynamic derivatives and maneuverability results of CFD and ASE methods 119
Table 3. Geometric constraints of submarine hull form optimization 122
Table 4. Pseudo-code of deep determinstic policy gradient algorithm for stealth design optimization 130
Table 5. Hyperparameters of DDPG 140
Table 6. Comparison of composition of reward components according to optimization cases 143
Table 7. Performance comparison of target strength reduction optimized hull form 158
Table 8. Compariso ns of flow noise analysis methods for underwater propellers 160
Table 9. Comparison of characteristics of underwater and aerial propellers 190
Table 10. Comparisons of radiation efficiency of aircraft and underwater propeller noise sources according to source order 190
Table 11. Noise contribution by source type of underwater propeller 191
Table 12. Main particulars of submarine and propeller 203
Table 13. Grid systems for convergence test 208
Table 14. Numerical settings for CFD calculation 209
Table 15. Self-propulsion points by model test and CFD analysis 215
Table 16. Comparison of sound pressure level of tonal noise 222
Table 17. CFD results for deriving optimal length-to-height ratio of front VRD 271
Table 18. CFD results for deriving minimum required length of front VRD 272
Table 19. CFD results for deriving optimal length-to-height ratio of rear VRD 280
Table 20. CFD results for deriving optimal length of rear VRD 281
Table 21. Comparisons of studies on improving submarine stealth performance 284
Table 22. Comparison of objective and reward composition according to optimization type 289
Table 23. Results of optimal dimensions of front and rear VRDs 296
Table 24. Comparison of hydrodynamic performance of candidate hull forms in self-propelled condition 302
Table 25. Comprehensive review of the performance of stealth candidate hull forms 326
Figure 1. Thesis structure and contents 35
Figure 2. Features of hull form of Virginia-class submarine 38
Figure 3. Development of Virginia-class fillet using composite materials 39
Figure 4. Features of hull form of Soryu-class submarine 41
Figure 5. Comparison of sail shape of Soryu-class and previous class 41
Figure 6. Comparison of hull form of Soryu-class and previous class 42
Figure 7. Features of hull form of Scorpéne-class submarine 43
Figure 8. Features of hull form of 212-class submarine 45
Figure 9. Features of hull form of 212CD-class submarine 46
Figure 10. Features of hull form of Suffren-class submarine 47
Figure 11. Features of hull form of Blekinge-class submarine 48
Figure 12. Features of hull form of Type 039C-class submarine 49
Figure 13. Concept of sonar cross section 56
Figure 14. Comparison of TS for cylinder at monostatic condition 56
Figure 15. Comparison of TS for rectangular plate at monostatic condition 57
Figure 16. Invisible surfacc elimination for submarine by using backface culling 62
Figure 17. Comparison of TS by invisible surface removal method 63
Figure 18. Comparison of TS by multiple reflection effect 63
Figure 19. Comparison of contributions to target strength of appendages and hull of SUBOFF 64
Figure 20. Comparison of contributions to target strength of appendages and hull of BB2 64
Figure 21. Comparison of time- and frequency-domain method for acoustic analogy 80
Figure 22. Helical coordinate system of propeller blade section 80
Figure 23. Logic comparison of machine learning types 81
Figure 24. Schematic diagram of reinforcement learning 84
Figure 25. Schematic diagram of stealth design optimization 96
Figure 26. Geometry of BeTSSi model 100
Figure 27. Dimensions of BeTSSi model 100
Figure 28. Geometric parameterization of submarine sail 101
Figure 29. Geometric parameterization of submarine rudder 101
Figure 30. Geometric parameterization of submarine hull section 102
Figure 31. Target strength analysis procedure 105
Figure 32. Receiver settings for target strength analysis 105
Figure 33. Validations of target strength calculation for BeTSSi model (solid line) reference result (dashed line) present result 106
Figure 34. Definition of underwater vehicle coordinate system and notation of hydrodynamic force and moment 117
Figure 35. Settings and result of CFD analysis for maneuverability verification 118
Figure 36. Maneuverability analysis procedure using ASE method 119
Figure 37. Structure of DDPG algorithm 127
Figure 38. Schematic diagram of TS reduction design optimization using deep reinforcement learning 138
Figure 39. Convergence trend of reward for optimization A 143
Figure 40. Convergence trend of reward for optimization B 144
Figure 41. Convergence trend of reward for optimization C 144
Figure 42. Comparison of learning trends according to geometric constraint reward 147
Figure 43. Comparison of reward shape according to maneuverability consideration 148
Figure 44. Comparison of target strength learning trends according to maneuverability reward 149
Figure 45. Comparison of maneuverability learning trends according to maneuverability reward 150
Figure 46. Target strength reduction hull form of optimization A 154
Figure 47. Target strength reduction hull form of optimization B 155
Figure 48. Target strength reduction hull form of optimization C 156
Figure 49. Comparison of target strength reduction hull form 157
Figure 50. Schematics of horseshoe vortex system around appendage 164
Figure 51. Schematic of the pirmary and secondary vortex in HSV system 164
Figure 52. Numerical simulation result using RANS for HSV generated from hydrofoil 167
Figure 53. Numerical simulation result using DES for HSV generated from hydrofoil at U=9 m/s, Re=5. 9 * E5 167
Figure 54. Numerical simulation result using DES for HSV generated from hydrofoil at U=15 m/s, Re=9. 8 * E5 168
Figure 55. Noise measurement of pusher propellers according to front wing configurations 171
Figure 56. Experiment on the effect of appendage vortex on propeller noise 171
Figure 57. Grid geometry and flow analysis results of the upper part of submarine 174
Figure 58. Self-noise analysis result of HSV using porous FW-H 175
Figure 59. Flow analysis results for the vortex generated around the sail and rudder of SUBOFF 177
Figure 60. Development process of vortex around SUBOFF hull form 178
Figure 61. Flow analysis results for the vortex generated around the sail of BB2 181
Figure 62. Development process of vortex around BB2 hull form 181
Figure 63. Vorticity analysis along streamlines around the BB2 appendages 182
Figure 64. Analysis results and schematic diagram of BB2 appendage vortex flowing into the propeller 182
Figure 65. Generation of unsteady thickness noise due to fluctuation of inflow velocity with time 191
Figure 66. Schematic diagrams of experimental set-up (top) experimental arrangement (bottom) three-cycle flow generator 195
Figure 67. Experimental results of tonal noise directivity for DTRC 3714 at three-cycle flow 195
Figure 68. Configurations of propeller in three-cycle inflow for validation case 196
Figure 69. Axial velocity distribution of three-cycle inflow of validation case 196
Figure 70. Comparison of BPF noise prediction for marine propeller by existing and developed noise model with experiments (top) 1st BPF (bottom)...[이미지참조] 197
Figure 71. Procedure of frequency-domain noise model 200
Figure 72. Unsteady noise sources of propeller blade induced by non-uniform wake 201
Figure 73. Geomtries of submarine and propeller (left) DARPA SUBOFF (right) INSEAN E1619 203
Figure 74. Configurations of experiment (top) submarine placement in the LCT test section (bottom) schematic of placement of hydrophones in the test section 204
Figure 75. Configurations of computational domain for submarine in LCT test section 208
Figure 76. Configurations of grid structure around the submarine (top) resistance model (bottom) self-propelled model 209
Figure 77. Grid geometry around the propeller of self-propelled model 210
Figure 78. Grid convergence test for resistance model 210
Figure 79. Comparison of total resistances from CFD analysis and experiments 211
Figure 80. Comparison of pressure coefficient from CFD analysis and experiments 211
Figure 81. Comparison of nominal wake distribution at x∕Loα=0.97 212
Figure 82. Comparison of CFD analysis results of pressure and velocity distributions for SUBOFF model at self-propelled condition 215
Figure 83. Vortex distribution of propeller inflow of SUBOFF model at self-propelled condition 216
Figure 84. Velocity distribution of propeller inflow of SUBOFF model at self-propelled condition 216
Figure 85. Unsteady noise sources of SUBOFF model at self-propelled condition 220
Figure 86. Propeller tonal noise in the first BPF of SUBOFF model using frequency domain noise model 221
Figure 87. Propeller loading and thickness noise in the first BPF of SUBOFF model using frequency domain noise model 221
Figure 88. Geometric models of Vortex reduction device s (top) convex type (bottom) concave type 226
Figure 89. Placement of VRDs around appendages (top) sail (bottom) rudder 227
Figure 90. Comparisons of vortex reductioin performance according to rear VRD type 228
Figure 91. Pressure distribution around sail and vorticity along the streamline (top) submarine without VRDs (bottom) submarine with VRDs 232
Figure 92. Pressure distribution around rudder and vorticity along streamline (top) submarine without VRDs (bottom) submarine with VRDs 233
Figure 93. Vorticity distribution at x/Loa=0.967 (left)submarine without VRDs (right) submarine with VRDs[이미지참조] 234
Figure 94. Axial velocity distribution at x/Loa=0.967 (top) submarine without VRDs (bottom) submarine with VRDs[이미지참조] 235
Figure 95. Unsteady noise sources of the reference blade (left) submarine without VRD s (right) submarine with VRDs 239
Figure 96. Directivity comparison of overall tone noise of submarine in 1st BPF 240
Figure 97. Directivity comparison of tone noise of submarine in 1st BPF (left) loading noise (right) thickness noise 240
Figure 98. Directivity comparison of overall tone noise of submarine in 2nd BPF 241
Figure 99. Directivity comparison of tone noise of submarine in 2nd BPF (left) loading noise (right) thickness noise 241
Figure 100. Two types of radius range at propeller inflow distribution 246
Figure 101. Determination of dominant vertex flow through wake analysis in two types of radius range 246
Figure 102. Comparison of 1st BPF noise contribution of unsteady source according to harmonic number and blade radius 247
Figure 103. Comparison of 2nd BPF noise contribution of unsteady source according to harmonic number and blade radius 248
Figure 104. Application priority according to reduction target noise 249
Figure 105. Schematic diagram of HSV and RV observation planes 253
Figure 106. Comparison of nondimensional vorticity of HSV generated from hydrofoils of different dimensions (top) hydr ofoil with maximum thickness of... 256
Figure 107. Comparison of nondimensional vorticity of RV generated from hydrofoils of different dimensions (top) hydrofoil with maximum thickness of... 257
Figure 108. Definition of installation location and dimensions of front VRD 261
Figure 109. Definition of installation location and dimensions of rear VRD 261
Figure 110. Definition of main curves of front VRD geometry 262
Figure 111. Definition of main curves of rear VRD geometry 262
Figure 112. CFD analysis domain and grid for VRD optimization 266
Figure 113. CFD analysis results for front VRD (Lt=0. 20 m, H=0. 20 m)[이미지참조] 267
Figure 114. CFD analysis results for front VRD (Lt=0. 40 m, H=0. 20 m)[이미지참조] 268
Figure 115. Required minimum length of front VRD depending on appendage thickness and Reynolds number 273
Figure 116. CFD analysis results for rear VRD (Leff=0. 05 m, H=0. 05 m)[이미지참조] 276
Figure 117. CFD analysis results for rear VRD (Lt=0. 10 m, H=0. 10 m)[이미지참조] 277
Figure 118. Optimal length of rear VRD depending on appendage thickness and Reynolds number 282
Figure 119. Procedures of stealth hull form design system 286
Figure 120. Convergence trend of reward for Stealth-A optimization 289
Figure 121. Convergence trend of reward for Stealth-B optimization 290
Figure 122. Convergence trend of reward for Stealth-C optimization 290
Figure 123. Stealth candidate hull form derived by Stealth-A optimization 293
Figure 124. Stealth candidate hull form derived by Stealth-B optimization 293
Figure 125. Stealth candidate hull form derived by Stealth-C optimization 294
Figure 126. Comparison of stealth candidate hull form 294
Figure 127. Combination of optimal front and rear VRD and Stealth-A hull form 297
Figure 128. Combination of optimal front and rear VRD and Stealth-B hull form 298
Figure 129. Combination of optimal front and rear VRD and Stealth-C hull form 299
Figure 130. Pressure distribution and vorticity along streamline for Base model hull form (top) sail (bottom) rudder 303
Figure 131. Pressure distribution and vorticity along streamline for Stealth-A hull form (top) sail (bottom) rudder 304
Figure 132. Pressure distribution and vorticity along streamline for Stealth-B hull form (top) sail (bottom) rudder 305
Figure 133. Pressure distribution and vorticity along streamline for Stealth-C hull form (top) sail (bottom) rudder 306
Figure 134. Comparison of isosurface(value=0. 2/s²) of Q-criterion around the sail 310
Figure 135. Comparison of isosurface(value=0. 2/s²) of Q-criterion around the rudder 311
Figure 136. Comparison of propeller inflow at x/Loa=0. 99 (left) vorticity distribution (right) velocity distribution[이미지참조] 312
Figure 137. Comparison of flow noise level at 1st BPF of stealth candidate hull forms (R=100 m)[이미지참조] 316
Figure 138. Noise source contribution analysis of base model 317
Figure 139. Noise source contribution analysis of Stealth-A hull form 317
Figure 140. Noise source contribution analysis of Stealth-B hull form 318
Figure 141. Noise source contribution analysis of Stealth-C hull form 318
Figure 142. TS performance of base model hull form(f=5 kHz, R=2 km) 322
Figure 143. TS performance of Stealth-A hull form (f=5 kHz, R=2 km) 322
Figure 144. TS performance of Stealth-B hull form (f=5 kHz, R=2 km) 323
Figure 145. TS performance of Stealth-C hull form (f=5 kHz, R=2 km) 323
Figure 146. TS contribution of basemodel hull form (azimuth/elev=90°/0°) 324
Figure 147. TS contribution of Stealth-A hull form (azimuth/elev=90°/0°) 324
Figure 148. TS contribution of Stealth-B hull form (azimuth/elev=90°/0°) 324
Figure 149. TS contribution of Stealth-C hull form (azimuth/elev=90°/0°) 324
Figure 150. Comparison of TS performance for stealth candidate hull forms (f=5 kHz, R=2 km, elevation angle=0°) 325