Digital Twin Method of Stress Field of Deep Submersible Spherical Shell Based on Simulation Database

doi:10.16182/j.issn1004731x.joss.24-0173

Abstract

Abstract:

This paper presents a method for predicting the stress field of deep diving spherical shells based on simulation databases and digital twin technology. By establishing simulation databases of stress field distribution of pressure-resistant spherical shells under different scales and loads, virtual sensing monitoring of stress states in other parts of the vessel is realized through finite sensor layout of pressure-resistant shells on the submersible. Based on the DT(digital twin) technology, a three-level virtual structure layer is constructed. The Level-1 DT layer realizes the spatial mapping and cloud image display from the finite element simulation model to the digital model. The error between the experimental and numerical results of the ultimate bearing capacity of the spherical shell is less than 9.4%. The Level-2 DT layer realizes the data sample deduction of the digital model by create database. The stress field distribution of the spherical shell under the condition that the size and load are not obtained in the simulation database is obtained by the local Lagrange interpolation method. The relative error of the stress interpolation result is 4.8%. The Level-3 DT layer develops a machine learning prediction function for the stress field distribution in the dangerous area of the deep-submersible spherical shell digital model. The BP neural network optimized by the particle swarm optimization algorithm ensures that the error between the prediction result and the simulation result is less than 1%. This method comprehensively considers the material properties, structural dimensions and environmental loads, which can provide a reference for the real-time safety assessment of the pressure hull structure, and realize the dynamic perception, intelligent diagnosis and scientific prediction of the dynamic stress field distribution of all deep-submersible spherical shells on the hull.

Key words: simulation database, digital twin, deep submerged spherical shell, stress field distribution, optimization algorithm

CLC Number:

TP391.9

Cao Yu, Li Jie, Wang Fang, Liu Zhixiang, Wang Xueliang. Digital Twin Method of Stress Field of Deep Submersible Spherical Shell Based on Simulation Database[J]. Journal of System Simulation, 2024, 36(8): 1764-1779.

Figures/Tables 31

Fig. 1

Fig. 2

Fig. 3

Fig. 4

Fig. 5

Table 1

Parameters of deep-sea pressure-resistant spherical shell

参数	值
耐压壳内径 $r / m m$	250
初始壁厚 $w / m m$	9
舱口盖外倾角 $s$ /(°)	45
弹性模量 $E / (k g ⋅ m m - 2)$	1.08×10⁵
泊松比 $v$	0.3
密度 $ρ / (k g ⋅ m m - 3)$	4.43×10^-9
材料	Ti-6Al-4V

Table 1

Fig. 6

Fig. 7

Fig. 8

Fig. 9

Fig. 10

Fig. 11

Fig. 12

Fig. 13

Fig. 14

Fig. 15

Fig. 16

Fig. 17

Table 2

Table 3

Tab.4

Fig. 18

Table 5

Fig. 19

Fig. 20

Fig. 21

Fig. 23

Fig. 22

Fig. 24

Table 6

Table 7

References 19

1	Wang Fang, Zhang Xuezhong, Kong Pinin, et al. An Improved Small-time-scale Crack Growth Rate Model Considering Overloading and Load-sustaining Effects for Deep-sea Pressure Hulls[J]. Ocean Engineering, 2022, 247: 110361.
2	Huang Xing, Wang Shaohua, Liu Quansheng, et al. Development of a Real-time Monitoring and Calculation Method for TBM Disc-cutter's Cutting Force in Complex Ground[J]. Geotechnical Testing Journal, 2022, 45(5): 961-984.
3	Chen Mingsheng, Li Chunbao, Han Zhaolong, et al. A Simulation Technique for Monitoring the Real-time Stress Responses of Various Mooring Configurations for Offshore Floating Wind Turbines[J]. Ocean Engineering, 2023, 278: 114366.
4	张涛, 胡嘉骏, 汪雪良, 等. 基于人工智能的三维结构全局应力求解方法研究[J]. 船舶力学, 2023, 27(2): 238-249.
	Zhang Tao, Hu Jiajun, Wang Xueliang, et al. Solving Approach for Global Stress Field of the 3D Structures Based on Artificial Intelligence[J]. Journal of Ship Mechanics, 2023, 27(2): 238-249.
5	Yao Ji, Wang Xueliang, Ye Cong, et al. Long-term Life Stress Mapping Algorithm of the Deep-sea Pressurized Spherical Shell Based on Digital-twin Technology[J]. Ocean Engineering, 2023, 286, Part 2: 115667.
6	Liu Xiaoben, Zheng Qian, Wu Kai, et al. Development of a Novel Approach for Strain Demand Prediction of Pipes at Fault Crossings on the Basis of Multi-layer Neural Network Driven by Strain Data[J]. Engineering Structures, 2020, 214: 110685.
7	Chen Jie, Liu Zhixin, Yin Zhengtong, et al. Predict the Effect of Meteorological Factors on Haze Using BP Neural Network[J]. Urban Climate, 2023, 51: 101630.
8	Huang Minjie, Jiang Jufu, Wang Ying, et al. Comparison and Error Evaluation of Arrhenius Model and Typical Machine Learning Algorithms for High-temperature Flow Stress Prediction of GH3536 Superalloy[J]. Materials Letters, 2023, 349: 134754.
9	毛子泉, 高家隆, 龚建兴, 等. 虚实结合仿真在军事领域的应用综述[J]. 系统仿真学报, 2023, 35(11): 2289-2311.
	Mao Ziquan, Gao Jialong, Gong Jianxing, et al. Application of Virtual-real Simulation in Military Field[J]. Journal of System Simulation, 2023, 35(11): 2289-2311.
10	熊群毓. 大数据时代MySQL数据库的应用分析[J]. 信息与电脑, 2023, 35(14): 209-212.
	Xiong Qunyu. Application Analysis of MySQL Database in the Era of Big Data[J]. Information & Computer, 2023, 35(14): 209-212.
11	胡沛然, 陈少辉. 权重归一化拉格朗日插值及其空间降尺度应用[J]. 遥感信息, 2019, 34(6): 63-71.
	Hu Peiran, Chen Shaohui. Weight Normalization Based Lagrange Interpolation and Its Application in Downscaling[J]. Remote Sensing Information, 2019, 34(6): 63-71.
12	Wu Yangwei, Li Aijun, Lei Su, et al. Prediction of Pyrolysis Product Yield of Medical Waste Based on BP Neural Network[J]. Process Safety and Environmental Protection, 2023, 176: 653-661.
13	米春, 李思颖, 牟佳祎, 等. 基于多传感器的城市隧道环境监测数据协同融合方法研究[J]. 现代隧道技术, 2023, 60(5): 177-185.
	Mi Chun, Li Siying, Mou Jiayi, et al. Study on Collaborative Fusion Method for Urban Tunnel Environmental Monitoring Data Based on Multi-sensor[J]. Modern Tunnelling Technology, 2023, 60(5): 177-185.
14	中国船级社. 潜水系统与潜水器入级规范: 标准编号缺失[EB/OL]. (2018-12-01) [2023-12-19]. .
	China Classification Society. Rules for Classification of Diving Systems and Submersibles[EB/OL]. (2018-12-01) [2023-12-19]. .
15	刘丹, 杜青海, 刘峰酉, 等. 基于神经网络的深海球形结构机器学习设计与优化[J]. 中国造船, 2022, 63(6): 161-173.
	Liu Dan, Du Qinghai, Liu Fengyou, et al. Machine Learning Design and Optimization of Deep-sea Spherical Shell Based on Artificial Neural Network[J]. Shipbuilding of China, 2022, 63(6): 161-173.
16	Wu Yu, Ding Jian, Wang Fang, et al. Research on the Quasi-static Collapse and Instantaneous Implosion of the Deep-sea Spherical Pressure Hull[J]. Marine Structures, 2022, 83: 103191.
17	张颖, 赖长亮, 和卫平, 等. 潜水器耐压壳结构选材应用综述[J]. 舰船科学技术, 2022, 44(5): 1-6.
	Zhang Ying, Lai Changliang, He Weiping, et al. Review of Materials Selection and Application of Submersible Pressure Hull Structure[J]. Ship Science and Technology, 2022, 44(5): 1-6.
18	刘汭琳, 满轲, 刘晓丽, 等. 基于GA-BP神经网络的富水砂层渣土改良效果预测[J]. 隧道建设(中英文), 2023, 43(增1): 222-232.
19	王娟, 杨小渝, 王宗国, 等. 基于混合学习算法的材料计算数据误差估计[J]. 系统仿真学报, 2018, 30(10): 3843-3852, 3860.
	Wang Juan, Yang Xiaoyu, Wang Zongguo, et al. Error Estimation for Material Simulation Data Based on Hybrid Learning Algorithm[J]. Journal of System Simulation, 2018, 30(10): 3843-3852, 3860.

插值点数量	应力/MPa	误差/‰
	87.63 (真实值)
3	87.877 (左二右一)	0.282
3	87.170 (左一右二)	0.523
4	87.523 (左二右二)	0.122
5	87.644 (左三右二)	0.316
5	86.786 (左二右三)	0.963
6	87.215 (左三右三)	0.474
7	88.016 (左四右三)	0.440
7	86.600 (左三右四)	1.175
8	87.308 (左四右四)	0.367

插值点数量	应力/MPa	误差/‰
	132.800 (真实值)
3	132.767 (左二右一)	0.248
3	132.867 (左一右二)	0.504
4	132.817 (左二右二)	0.128
5	132.820 (左三右二)	0.151
5	132.840 (左二右三)	0.301
6	132.830 (左三右三)	0.226
7	132.831 (左四右三)	0.233
7	132.782 (左三右四)	0.136
8	132.837 (左四右四)	0.279

节点	外压/Mpa	壁厚/mm	节点1	节点2	节点3	节点4	节点5	节点6	节点7	节点8
1	1	9	14.5	14.4	14.4	13.5	14.4	10.9	14.5	13.7
2	1	10	9.8	13.1	9.8	12.4	12.4	13.1	12.4	13.1
…	…	…	…	…	…	…	…	…	…	…
5 480	30	27.1	231.5	129.2	162.6	129.2	162.6	219.5	219.5	129.2
5 481	30	27.2	231.6	128.7	162.1	128.7	162.1	219.5	219.5	128.7
5 482	30	27.3	231.8	128.2	161.6	128.2	161.6	219.5	219.5	128.2
5 483	30	27.4	231.9	127.7	161.1	127.7	161.1	219.5	219.5	127.7
5 484	30	27.5	232.1	127.3	160.6	127.2	160.5	219.6	219.6	127.2
5 485	30	27.6	232.3	126.8	160.1	126.7	160.1	219.6	219.6	126.7
5 486	30	27.7	232.5	126.3	159.6	126.2	159.6	219.7	219.7	126.2
5 487	30	27.8	232.7	125.8	159.1	125.7	159.1	219.7	219.7	125.8
5 488	30	27.9	232.8	125.4	158.6	125.3	158.6	219.7	219.7	125.3
5 489	30	28.0	232.9	124.9	158.1	124.8	158.1	219.7	219.7	124.8
5 490	30	29.0	229.2	120.3	153.4	120.3	153.5	216.3	216.3	120.3
…	…	…	…	…	…	…	…	…	…	…
9 140	50	27.1	385.7	215.4	271.0	215.4	271.0	365.8	365.8	215.3
9 141	50	27.2	385.9	214.5	270.2	214.5	270.2	365.8	365.8	214.4
9 142	50	27.3	386.2	213.7	269.4	213.7	269.3	365.8	365.8	213.6
9 143	50	27.4	386.5	212.9	268.5	212.9	268.5	365.9	365.9	212.8
9 144	50	27.5	386.9	212.1	267.7	212.1	267.7	366.0	366.0	211.9
9 145	50	27.6	387.2	211.3	266.9	211.3	266.8	366.1	366.1	211.2
9 146	50	27.7	387.5	210.5	266.0	210.5	266.0	366.2	366.1	210.4
9 147	50	27.8	387.8	209.7	265.2	209.7	265.2	366.2	366.2	209.6
9 148	50	27.9	388.0	208.9	264.4	208.9	264.4	366.2	366.2	208.8
9 149	50	28.0	388.2	208.1	263.6	208.1	263.6	366.2	366.2	208.0
9 150	50	29.0	382.0	200.5	255.7	200.5	255.8	360.4	360.4	200.4

参数神经元数量	均方根误差	参数神经元数量	均方根误差
3	0.000 618 78	8	0.000 603 58
4	0.000 264 22	9	0.000 152 86
5	0.000 195 16	10	0.000 203 48
6	0.000 177 27	11	0.000 085 343
7	0.000 615 34	12	0.000 013 195

参数	6-4划分	7-3划分	8-2划分
训练集RMSE	2.47	2.38	2.35
训练集MAPE/%	2.65	2.49	1.99
测试集RMSE	19.69	15.42	5.96
测试集MAPE/%	3.07	2.25	1.07