力学进展

Advances in control of turbulence by artificial intelligence: Systems, algorithms, achievements and data analysis methods

WU Zhi, FAN De Wei, ZHOU Yu

2023, 53(2): 273-307. doi: 10.6052/1000-0992-22-045

Abstract(4375)

HTML (2456)

PDF(1476)

Abstract:
Turbulence control involves fluid dynamics and control theory, and is of great importance to many fields such as aeronautics and astronautics, vehicle, wind power generation, etc. Due to the complexity of turbulence, traditional control methods face many bottlenecks in the field of turbulence control. The development of artificial intelligence (AI) technology provides a tool to break through these bottlenecks. This paper briefly summarizes the applications of AI in turbulence control reported in the literature, focusing on AI control systems, algorithms, and the outstanding achievements achieved in different turbulence control applications, as well as the first attempt by the author's team to analyze the big data generated by the AI control system to discover important information and even the control scaling law. The challenges and future prospects are also analyzed.

Review on research progress of nonlinear vibration isolation and time-delayed suppression method

SUN Xiuting, QIAN Jiawei, QI Zhifeng, XU Jian

2023, 53(2): 308-356. doi: 10.6052/1000-0992-22-048

Abstract(2339)

HTML (770)

PDF(703)

Abstract:
Vibration problem is closely related to human daily life and the development of science and technology. Undersigned vibration would affect human health, also cause the failure of buildings, mechanical equipment and precision instruments. In the fields of aerospace, automotive engineering, marine engineering, large structure and precision instrument processing, elastic components or control devices are utilized to form effective vibration isolation environment. Effective vibration isolation can improve the safety, stability, controllability and comfort of human and equipment. However, the design method based on linear theory appears irreconcilable contradiction for applications. The stiffness of the vibration isolation structure reduces for wider band for vibration isolation, which results the decrease of loading capacity. In this paper, the high-static-low-dynamic design methods of nonlinear vibration isolators are reviewed in detail. Based on the dynamic modeling and analysis of nonlinearity induced by different mechanisms, the quasi-zero stiffness design criteria are presented, which provides the selection principle of the kind of isolator in aerospace, precision machining, automobile and ship. Then, attentions are paid on the design methods of bionics and metastructure for vibration suppression, which causes new problems and challenges in the explanation on mechanical bionic mechanism and regulation on the nonlinear restoring force constitutive. The studies on new methods, strategies and experimental techniques are introduced. Finally, faced to the development of light-weight and flexibility of structures, the vibration isolation effect is limited by the flexibility of parts. Even through coupling multi-layer quasi-zero-stiffness structures are assembled, it is difficult to realize the rapid positioning of the isolated structure in space. Considering the amplitude and frequency modulation by time delay, different design methods of vibration suppression with time delay are proposed. Successful cases of time-delayed suppression of flexible low-frequency vibration are given, which provides a possibility for effective, simple and rapid active vibration isolation and vibration elimination under extreme working conditions. In the future, based on the new analysis method according to data-driven, the nonlinear vibration isolation and suppression technology would realize more accurate and intelligent control effect in complex working conditions.

Stochastic resonance of multi-stable dynamical systems: A review

JIN Yanfei, XU Pengfei, LI Yongge, MA Jinzhong, XU Yong

2023, 53(2): 357-394. doi: 10.6052/1000-0992-22-047

Abstract(1609)

HTML (761)

PDF(463)

Abstract:
The nonlinear stochastic dynamical system has been an important subject in areas of mechanics, mathematics, engineering and so on, and finds various applications in different fields like mechanical engineering, aerospace engineering, ocean engineering, and biology. The multi-stable dynamical systems are conceptual nonlinear systems, coupling with stochastic excitations, which can exhibit complex dynamical behaviors, such as stochastic resonance and stochastic bifurcation. The stochastic resonance theory has been utilized effectively in many areas related to stochastic dynamics such as mechanical fault diagnosis, weak signal detection and vibration energy harvesting. This paper overviews the fundamental theories, methods and engineering applications of stochastic resonance in multi-stable dynamical systems. We introduce recent advances in theories and measure index of stochastic resonance via several classic examples of nonlinear dynamical systems. Then, we summarize the results of multi-stable dynamical systems under the excitation of different types of noise. The tri-stable and periodic systems are illustrated to show the occurrence principle, evolution mechanism and investigated techniques. Finally, three engineering applications of multi-stable dynamical systems are surveyed. Some open problems are presented to close this paper.

Ballistic performance of lightweight cellular sandwich structures: A review

CUI Tianning, QIN Qinghua

2023, 53(2): 395-432. doi: 10.6052/1000-0992-23-002

Abstract(1822)

HTML (506)

PDF(522)

Abstract:
The classical cellular sandwich panels composed of two thin, stiff face sheets separated by a novel cellular core are a class of promising advanced lightweight multi-functional structures, possessing high specific stiffness, high specific strength, excellent mitigation and energy absorption, and high designability. Cellular sandwich structures have been paid much attention in many fields, such as the aerospace industry, transportation and structural protection. Moreover, the success cases have been presented in practical engineering applications. In order to clarify the mechanisms of penetration and energy dissipation and extend the application ranges, investigations on ballistic performance of lightweight cellular sandwich structures are reviewed and prospected. Firstly, the structural features and types of lightweight cellular sandwich structures are summarized. Next, mechanisms of penetration and energy dissipation, and lightweight design are reviewed systematically. Finally, the problems and challenges existing in the current research on ballistic performance of lightweight cellular sandwich structures are prospected.

Prospects of multi-paradigm fusion methods for fluid mechanics research

ZHANG Weiwei, WANG Xu, KOU Jiaqing

2023, 53(2): 433-467. doi: 10.6052/1000-0992-22-050

Abstract(1996)

HTML (779)

PDF(605)

Abstract:
Experimental observation, theoretical research and numerical simulation are the basic research paradigms in many disciplines, including fluid mechanics. Since the 21st century, artificial intelligence based on big data has become an important driving force, leading to new scientific and technological revolution and industrial transformation. This is known as the data-intensive scientific research paradigm, which forms the fourth research paradigm. Similarly, data-driven machine learning has also become an emerging research direction in fluid mechanics and promoted the progress in intelligent fluid mechanics. However, compared to traditional data-intensive research paradigm that relies on "Internet and big data", research on intelligent fluid mechanics has its own unique background. For example, compared to high-dimensional flow state, geometric boundary conditions, and the inherent high-dimensional, cross-scale, random, and nonlinear characteristics of complex flow, research in data-driven fluid mechanics essentially handles large data but small samples. Although there are three major research paradigms in fluid mechanics, the integration among research paradigms is very low, where engineering optimization simply corrects data from multiple sources. Multi-source data fusion can alleviate several dilemmas, like small data sample from a single source, difficulties in modelling, and the insufficient utilization of low-fidelity data, it still fails to fully integrate theoretical models, expert knowledge and experience from the basic paradigms.Therefore, based on the fourth paradigm driven by artificial intelligence, the organic combination of three major research topics including experiment, theoretical model and numerical simulation, developing date and knowledge jointly driven multi-paradigm fusion methods for fluid mechanics, have become urgent to solve major practical engineering problems, as well as to satisfy the need for the development of the connotation and the characteristics of fluid mechanics in the new era.

Review of dynamics simulation methods for multi-field coupling systems

WANG Guoping, TAO Ling, RONG Bao, RUI Xiaoting

2023, 53(2): 468-495. doi: 10.6052/1000-0992-22-043

Abstract(2741)

HTML (954)

PDF(553)

Abstract:
Modern engineering system is usually a multi-field coupling system with complex structure/mechanical system as the main body, which integrates thermal, electromagnetic, fluid and other subsystems. The dynamics modeling of these systems are complex and their numerical calculate are also difficult, which bring unprecedented challenges to the efficient and accurate evaluation and design optimization of the system dynamic characteristics. In this paper, the research progress of multi-field coupling dynamics simulation methods for complex engineering systems are reviewed in detail, including the basic strategies of multi-field coupling dynamics modeling and numerical solution, mesh deformation processing method, coupling data exchange technology, numerical calculation efficiency, etc. On this basis, for uncertain multi-field coupling engineering systems, their uncertainty analysis and reliability evaluation methods under single and hybrid uncertainties are discussed in detail. This review paper is expected to provide some useful references for related researchers.

2023 Vol. 53, No. 2

NoticeMore >

Author CenterMore >

Related linksMore >