Nonlinear Aeroelastic Prediction in Transonic Buffeting Flow by Deep Neural Network

气动弹性 跨音速 空气动力学 翼型 计算流体力学 非线性系统 空气动力 计算机科学 控制理论(社会学) 机械 结构工程 物理 工程类 量子力学 人工智能 控制(管理)
作者
Zihao Dou,Chuanqiang Gao,Weiwei Zhang,Yang Tao
出处
期刊:AIAA Journal [American Institute of Aeronautics and Astronautics]
卷期号:61 (6): 2412-2429
标识
DOI:10.2514/1.j061946
摘要

Transonic buffet is an aerodynamic phenomenon of self-sustained shock oscillations. The aeroelastic problem caused by it is very complex, including two different dynamic modes: forced vibration and frequency lock-in. The vibration of the structure has a negative influence on the fatigue life of the aircraft. Especially in the region of frequency lock-in, the limit cycle oscillations occur due to the instability of the structural mode. Researchers have accurately predicted the region of frequency lock-in in transonic buffet and have clarified its mechanism by using a linear aerodynamic model. However, the nonlinear aeroelastic modeling and prediction of the transonic buffet remain to be solved. The long short-term memory (LSTM) deep neural network is suitable for predicting the time-delayed effects of unsteady aerodynamics. And it has achieved remarkable results in sequential data modeling. In the present work, a nonlinear model is developed for the aeroelastic system with NACA0012 airfoil in transonic buffeting flow and validated with the coupled computational fluid dynamics/computational structural dynamics (CFD/CSD) simulation. First, the data set and the loss function are specially designed. Then, the reduced-order model (ROM) based on the LSTM of the flow is built by using unsteady Reynolds-averaged Navier–Stokes computations data in a post-buffet state. By coupling the ROM and the single degree-of-freedom equation for the pitching angle, the nonlinear aeroelastic model is finally produced. The results show that the phenomenon of frequency lock-in and the self-sustained buffeting aerodynamics are precisely reconstructed. And the model has a strong generalization ability and can reproduce complex vibrations caused by competition between different modes. In short, the model can replace the CFD/CSD method in the current case with high efficiency and accuracy. The method can be used for modeling and prediction of other various complex aeroelastic systems.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
更新
大幅提高文件上传限制,最高150M (2024-4-1)

科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
wwho_O发布了新的文献求助30
19秒前
打打应助不散的和弦采纳,获得10
20秒前
Ting应助狒狒采纳,获得10
22秒前
why发布了新的文献求助10
23秒前
24秒前
liuarise发布了新的文献求助10
24秒前
lz完成签到,获得积分10
25秒前
调研昵称发布了新的文献求助10
29秒前
大个应助大巧若拙采纳,获得10
29秒前
29秒前
清川映叶应助科研通管家采纳,获得10
30秒前
小蘑菇应助科研通管家采纳,获得10
30秒前
30秒前
852应助科研通管家采纳,获得10
30秒前
慕青应助科研通管家采纳,获得10
30秒前
SIiveryyyy完成签到,获得积分10
30秒前
30秒前
JamesPei应助科研通管家采纳,获得10
30秒前
30秒前
31秒前
31秒前
32秒前
34秒前
Ting应助RNAPW采纳,获得10
36秒前
37秒前
hyl1115发布了新的文献求助10
38秒前
HBUTL发布了新的文献求助10
38秒前
kuangweiming完成签到,获得积分10
39秒前
隐形曼青应助陶醉书包采纳,获得10
1秒前
1秒前
小天发布了新的文献求助10
2秒前
顾矜应助笨笨小熊猫采纳,获得10
2秒前
曾哥帅发布了新的文献求助10
6秒前
6秒前
zzz发布了新的文献求助10
9秒前
12秒前
小天完成签到,获得积分10
14秒前
fenghuu发布了新的文献求助10
14秒前
执着易形发布了新的文献求助10
18秒前
18秒前
高分求助中
Production Logging: Theoretical and Interpretive Elements 2000
BIOLOGY OF NON-CHORDATES 1000
进口的时尚——14世纪东方丝绸与意大利艺术 Imported Fashion:Oriental Silks and Italian Arts in the 14th Century 800
Autoregulatory progressive resistance exercise: linear versus a velocity-based flexible model 550
The Collected Works of Jeremy Bentham: Rights, Representation, and Reform: Nonsense upon Stilts and Other Writings on the French Revolution 320
Med Surg Certification Review Book: 3 Practice Tests and CMSRN Study Guide for the Medical Surgical (RN-BC) Exam [5th Edition] 300
Play from birth to twelve: Contexts, perspectives, and meanings – 3rd Edition 300
热门求助领域 (近24小时)
化学 医学 生物 材料科学 工程类 有机化学 生物化学 物理 内科学 纳米技术 计算机科学 化学工程 复合材料 基因 遗传学 物理化学 催化作用 细胞生物学 免疫学 冶金
热门帖子
关注 科研通微信公众号,转发送积分 3358140
求助须知:如何正确求助?哪些是违规求助? 2981312
关于积分的说明 8698638
捐赠科研通 2662919
什么是DOI,文献DOI怎么找? 1458178
科研通“疑难数据库(出版商)”最低求助积分说明 675060
邀请新用户注册赠送积分活动 666078