Acta Mechanica Sinica
The Journal covers all disciplines in the field of theoretical and applied mechanics, including solid mechanics, fluid mechanics, dynamics and control, and biomechanics. It explores analytical, computational and experimental progresses in all areas of mechanics. The Journal also encourages research in interdisciplinary subjects, and servess...
CN 11-2063/O3
ISSN 0567-7718(Print)
ISSN 1614-3116(Online)
Editors-in-Chief:
Professor Xiaojing Zheng
Professor Xuesong Wu
Professor Zhigang Suo
Impact Factor: 1.975
Citation: Acta Mech. Sin.
Frequency: Bimonthly
Focus Articles

 
Editorial & News
 
Current Issue Online First Archive Springer-Pre Most Read Articles Most Downloaded
Acta Mechanica Sinica--2021, 37 (12)   Published: 25 December 2021
THEMED PAPERS ON ARTIFICIAL INTELLIGENCE IN FLUID MECHANICS
Artificial intelligence in fluid mechanics
Wei-Wei Zhang,Bernd R. Noack
Abstract

The rapid generation of high-quality flow data and the development of increasingly powerful artificial intelligence methods foster novel highly fruitful research paradigms for solving big challenge problems in fluid mechanics. This paradigm change marks the birth of a novel field of research—intelligent fluid mechanics (IFM).

Applying machine learning to study fluid mechanics
Steven Brunton
Abstract
This paper provides a short overview of how to use machine learning to build data-driven models in fluid mechanics. The process of machine learning is broken down into five stages: (1) formulating a problem to model, (2) collecting and curating training data to inform the model, (3) choosing an architecture with which to represent the model, (4) designing a loss function to assess the performance of the model, and (5) selecting and implementing an optimization algorithm to train the model. At ea...
Physics-informed neural networks (PINNs) for fluid mechanics: a review
Shengze Cai, Zhiping Mao, Zhicheng Wang, Minglang Yin, George Em Karniadakis
Abstract
Despite the significant progress over the last 50 years in simulating flow problems using numerical discretization of the Navier-Stokes equations (NSE), we still cannot incorporate seamlessly noisy data into existing algorithms, mesh-generation is complex, and we cannot tackle high-dimensional problems governed by parametrized NSE. Moreover, solving inverse flow problems is often prohibitively expensive and requires complex and expensive formulations and new computer codes. Here, we review flow...
Genetic-algorithm-based artificial intelligence control of a turbulent boundary layer
Jianing Yu, Dewei Fan, Bernd. R. Noack, Yu Zhou
Abstract
An artificial intelligence (AI) open-loop control system is developed to manipulate a turbulent boundary layer (TBL) over a flat plate, with a view to reducing friction drag. The system comprises six synthetic jets, two wall-wire sensors, and genetic algorithm (GA) for the unsupervised learning of optimal control law. Each of the synthetic jets through rectangular streamwise slits can be independently controlled in terms of its exit velocity, frequency and actuation phase. Experiments are conduc...
Practical framework for data-driven RANS modeling with data augmentation
Xianwen Guo, Zhenhua Xia, Shiyi Chen
Abstract
Inspired by the iterative procedure of computing mean fields with known Reynolds stresses (Theoret. Appl. Mech. Lett., 2021), we proposed a way to achieve data augmentation by utilizing the intermediate mean fields after proper selections. We also proposed modifications to the Tensor Basis Neural Network (J. Fluid Mech., 2016) model. With the modification of the learning targets and the inclusions of wall distance and logarithm of normalized eddy viscosity in the model inputs, the modified versi...
Multilayer perceptron neural network activated by adaptive Gaussian radial basis function and its application to predict lid-driven cavity flow
Qinghua Jiang, Lailai Zhu, Chang Shu, Vinothkumar Sekar
Abstract
To improve the performance of multilayer perceptron (MLP) neural networks activated by conventional activation functions, this paper presents a new MLP activated by univariate Gaussian radial basis functions (RBFs) with adaptive centers and widths, which is composed of more than one hidden layer. In the hidden layer of the RBF-activated MLP network (MLP-RBF), the outputs of the preceding layer are first linearly transformed and then fed into the univariate Gaussian RBF, which exploits the highly...
Deconvolutional artificial-neural-network framework for subfilter-scale models of compressible turbulence
Zelong Yuan, Yunpeng Wang, Chenyue Xie, Jianchun Wang
Abstract
We establish a deconvolutional artificial-neural-network (D-ANN) approach in large-eddy simulation (LES) of compressible turbulent flow. Filtered variables in the neighboring locations are taken as the inputs of D-ANN to recover original (unfiltered) variables, including density, momentum and pressure. The scale-similarity form is adopted to reconstruct subfilter-scale (SFS) terms. The proposed D-ANN models can give better a priori predictions of the sub-filter stress and heat flux than the clas...
Bayesian optimization for active flow control
Antoine B. Blanchard, Guy Y. Cornejo Maceda, Dewei Fan, Yiqing Li, Yu Zhou, Bernd R. Noack, Themistoklis P. Sapsis
Abstract
A key question in flow control is that of the design of optimal controllers when the control space is high-dimensional and the experimental or computational budget is limited. We address this formidable challenge using a particular flavor of machine learning and present the first application of Bayesian optimization to the design of open-loop controllers for fluid flows. We consider a range of acquisition functions, including the recently introduced output-informed criteria of Blanchard and Saps...
Aerodynamic modeling using an end-to-end learning attitude dynamics network for flight control
Tun Zhao, Gong Chen, Xiao Wang, Enmi Yong, Weiqi Qian
Abstract
A novel identification method of aerodynamic models using a physics neural network, named the attitude dynamics network, which incorporates the attitude dynamics of an aircraft without any prior aerodynamic knowledge, is proposed. Then a learning controller, which combines feedback linearization with sliding mode control, is developed by introducing the learned aerodynamic models. The merit of the identification method is that the aerodynamic models can be learned end-to-end by the physics netwo...
Data-driven bayesian inference of turbulence model closure coefficients incorporating epistemic uncertainty
Daigo Maruyama, Philipp Bekemeyer, Stefan G?rtz, Simon Coggon, Sanjiv Sharma
Abstract
In computational fluid dynamics (CFD), various turbulence models are widely used to numerically investigate turbulent flows in academic and industrial applications. The closure coefficients in the models are typically seen as universal constants that were calibrated to match experimental results of representative, simplified flow experiments. In this article, we introduce a framework for statistical inference of the closure coefficients using machine learning methods. The objective of this frame...
 
v
Copyright © Acta Mechanica Sinica Editorial Office
Address: No.15 Bei Si Huan Xi Road, Beijing 100190, China
Telephone: +86-(0)10-62536271   Fax:+86-(0)10-82543907
E-mail: actams@cstam.org.cn
Technical support: Beijing Magtech Co.ltd support@magtech.com.cn