Article
Mechanics
Miguel P. Encinar, Javier Jimenez
Summary: The algorithm introduced by Jimenez (J. Fluid Mech., vol. 854, 2018, R1) is used to identify the flow patterns of causal significance in three-dimensional isotropic turbulence. The study finds that the dimensions of the perturbations introduced in the flow are controlled by the kinetic energy content and the enstrophy and dissipation, and affect their significance in the flow. Strain is found to be more efficient than vorticity in propagating the perturbation contents to other regions of the flow. The findings suggest that manipulating strain-dominated vortex clusters is more effective in controlling turbulent flows.
JOURNAL OF FLUID MECHANICS
(2023)
Article
Mechanics
Jie Ren, Xuerui Mao, Song Fu
Summary: The research introduces an image-based flow decomposition method, utilizing EWT modes to represent fluid physics with different scales. This approach shows promise in decomposing instantaneous three-dimensional flows and effectively extracting structural features and instabilities within the flow.
JOURNAL OF FLUID MECHANICS
(2021)
Article
Mechanics
Te-Yao Chiu, Chien-Chou Tseng, Chien-Cheng Chang, Yi-Ju Chou
Summary: This study investigated the influence of coherent structures on the aerodynamic forces acting on an aerofoil. The results showed that these structures have a significant impact on drag and lift forces, and can be identified and quantified using the SPOD algorithm and vorticity force analysis.
JOURNAL OF FLUID MECHANICS
(2023)
Article
Mechanics
J. G. Chen, J. C. Vassilicos
Summary: A theory of non-homogeneous turbulence is developed and applied to boundary-free shear flows. The theory introduces assumptions of inner and outer similarity for the non-homogeneity of two-point statistics, and predicts power-law scalings of second-order structure functions that have some similarities with but also some differences from Kolmogorov scalings. Comparisons with structure function data from three qualitatively different turbulent wakes provide support for the theory's predictions but also raise new questions for future research.
JOURNAL OF FLUID MECHANICS
(2022)
Article
Mechanics
J. G. Chen, C. Cuvier, J. -m. Foucaut, Y. Ostovan, J. C. Vassilicos
Summary: The study experimentally investigated the scaling of turbulent energy dissipation rate in the cross-stream direction of turbulent wake flows generated by two side-by-side square prisms. The results showed that the normalised turbulence dissipation coefficient for residual incoherent turbulence scaled similarly along the cross-stream direction after removal of large-scale coherent structures.
JOURNAL OF FLUID MECHANICS
(2021)
Article
Mechanics
Yixiao Shao, Jie Su, Yu Tu, Limin Kuang, Zhaolong Han, Kai Zhang, Dai Zhou
Summary: The actuator line model is used to study the wind turbine cluster composed of horizontal- and vertical-axis wind turbines in a vertically staggered layout. The results show that placing a vertical-axis wind turbine (VAWT) upwind of a horizontal-axis wind turbine (HAWT) can enhance the power generation of the HAWT by about 100 kW. Placing the VAWT downstream of the HAWT slightly reduces the power generation efficiency of the HAWT but increases the total power by about 60 kW. Placing the VAWT between the two HAWTs does not have a significant beneficial effect on the downstream HAWT, but the wind turbine cluster still generates 50 kW more power than without the VAWT. Overall, the collocation of VAWTs can effectively utilize otherwise wasted wind resources and increase the power generation density of wind farms.
Article
Mechanics
Tomoaki Watanabe, Yulin Zheng, Koji Nagata
Summary: This study investigates the decay of stably stratified turbulence generated by a towed rake of vertical plates through direct numerical simulations. The results are compared with the theory of stably stratified axisymmetric Saffman turbulence. It is found that under certain conditions, the decay of various quantities follows the power laws predicted for low buoyancy Reynolds number Saffman turbulence, but in some cases, the decay process no longer follows power laws.
JOURNAL OF FLUID MECHANICS
(2022)
Article
Mechanics
Guodan Dong, Jianhua Qin, Zhaobin Li, Xiaolei Yang
Summary: In this study, the characteristics of wind turbine wakes for three different blade designs were investigated using large-eddy simulations with the actuator surface model. The results show that the blade designs influence the velocity deficit, turbulence kinetic energy, and wake meandering. The NREL-Root design exhibits higher velocity deficit, the NREL-Tip design has higher turbulence kinetic energy in the near wake, and the NREL-Root design has higher turbulence kinetic energy in the far wake.
JOURNAL OF FLUID MECHANICS
(2023)
Article
Mechanics
Jiabin Wang, Guglielmo Minelli, Gioacchino Cafiero, Gaetano Iuso, Kan He, Branislav Basara, Guangjun Gao, Sinisa Krajnovic
Summary: This paper presents a numerical investigation of the effects of the moving ground and rotating wheels on the turbulent flow around a 1/10 scaled square-back van model. The study compares the performance of partially averaged Navier-Stokes (PANS), large eddy simulation (LES), and particle image velocimetry (PIV) methods in predicting the flow characteristics. The results show that PANS accurately predicts the flow field, even with a low-resolution grid, providing good guidance for industrial research in predicting the turbulent flow around the square-back van model.
JOURNAL OF FLUID MECHANICS
(2023)
Article
Mechanics
Jianyun Yangzhou, Jiafeng Wu, Zhaokai Ma, Xun Huang
Summary: In this study, the aeroacoustic sources of a two-bladed propeller ingesting an aerofoil wake were investigated using large eddy simulation and two different source identifying approaches. The numerical beamforming approach determined the phase variations of sources at low to mid frequencies and revealed that high-frequency sources are phase-independent. A new near-field aeroacoustic source analysis approach based on the acoustic analogy was developed to improve the spatial resolution of source identification. The proposed analysis approaches extended the capability of computational fluid dynamics and enabled the detailed study of noise generation mechanisms of wake-ingesting propeller noise.
JOURNAL OF FLUID MECHANICS
(2023)
Article
Mechanics
J. L. Ortiz-Tarin, S. Nidhan, S. Sarkar
Summary: The characteristics of the wake of a slender body with a turbulent boundary layer at high Reynolds numbers are studied, showing a faster decay rate once complete self-similarity is achieved, which differs from the classic high Reynolds number decay law.
JOURNAL OF FLUID MECHANICS
(2021)
Article
Mechanics
Junye Wang, Kan Wang, Meng Wang
Summary: Large-eddy simulation combined with the Ffowcs Williams-Hawkings equation was used to study noise generation by a rotor ingesting the turbulent wake of a circular cylinder. Results showed good agreement between computed sound pressure levels and experimental measurements across a wide range of frequencies. Rotor in thrusting condition produced stronger sound compared to zero thrust condition, and effects of rotor on wake turbulence were found to be relatively small.
JOURNAL OF FLUID MECHANICS
(2021)
Article
Mechanics
Antonio Posa, Riccardo Broglia, Mario Felli, Marta Cianferra, Vincenzo Armenio
Summary: The acoustic signature of a seven-bladed submarine propeller is characterized using the acoustic analogy method. Results show that the nonlinear terms of the governing equation dominate away from the propeller, while the linear terms decay downstream. The propeller's acoustic signature is mainly tonal in the near field, due to the thickness and loading components of noise from the surface of the propeller and the periodic perturbation caused by its tip vortices. The faster development of instability in the tip vortices compared to the hub vortex leads to an energy cascade towards higher frequencies, contributing to broadband noise.
JOURNAL OF FLUID MECHANICS
(2022)
Article
Mechanics
Vincent Mons, Olivier Marquet
Summary: The RANS-based data assimilation plays a vital role in data-driven approaches by enhancing experimental data and identifying turbulence model corrections. In this study, a sensor placement methodology is developed to infer the full mean flow from a few punctual mean velocity measurements. Two objectives are targeted: the correct reconstruction of dominant singular modes of the linearized RANS equations and the identification of model corrections using a second-order adjoint-based approach to optimize sensor locations and ensure the effectiveness of the inverse problem.
JOURNAL OF FLUID MECHANICS
(2021)
Article
Mechanics
Antonio Posa
Summary: This study investigates the tip vortices shed by two marine propellers using large-eddy simulation and a cylindrical grid with 5 billion points. It compares a tip-loaded design with winglets against a conventional one at specific operating conditions. The results show that the tip-loaded propeller achieves improved performance but also produces more intense tip vortices. These vortices originate from the edge of each winglet and the junction between the winglets and the blades and merge downstream. The simulations highlight the importance of carefully optimizing the geometry of the winglets to achieve weaker vortices and lower negative pressure peaks at their core.
JOURNAL OF FLUID MECHANICS
(2022)
Article
Multidisciplinary Sciences
Taichi Nakamura, Kai Fukami, Koji Fukagata
Summary: This paper investigates the fundamental differences between neural networks and linear stochastic estimation in fluid-flow regressions. Through comparisons and analyses of two fluid-flow problems, the study demonstrates that neural networks outperform linear methods due to the presence of nonlinear activation functions.
SCIENTIFIC REPORTS
(2022)
Article
Engineering, Mechanical
Kai Fukami, Byungjin An, Motohiko Nohmi, Masashi Obuchi, Kunihiko Taira
Summary: This paper presents a machine learning technique that accurately estimates the state of turbulent flow in engineering systems using limited sensor measurements. The technique can reconstruct turbulent vortical structures in a pump sump from sparse surface pressure measurements and accurately estimate flow with only a few sensor measurements, identifying the presence of adverse vortices.
JOURNAL OF FLUIDS ENGINEERING-TRANSACTIONS OF THE ASME
(2022)
Article
Mathematics, Applied
Masaki Morimoto, Kai Fukami, Romit Maulik, Ricardo Vinuesa, Koji Fukagata
Summary: The paper utilizes Gaussian stochastic weight averaging (SWAG) to assess the epistemic uncertainty in neural-network-based function approximation for fluid flows. With SWAG, multiple models with different combinations of weights can be created to obtain ensemble predictions. The average of the ensemble represents the mean estimation, while the standard deviation can be used to construct confidence intervals for uncertainty quantification. The method is applicable for various complex datasets and network architectures. The authors demonstrate its applicability for different types of neural networks and find that SWAG provides physically-interpretable confidence-interval estimates.
PHYSICA D-NONLINEAR PHENOMENA
(2022)
Article
Mechanics
J. H. Marques Ribeiro, Chi-An Yeh, Kunihiko Taira
JOURNAL OF FLUID MECHANICS
(2023)
Article
Thermodynamics
Marco Atzori, Fermin Mallor, Ramon Pozuelo, Koji Fukagata, Ricardo Vinuesa, Philipp Schlatter
Summary: For adverse-pressure-gradient turbulent boundary layers, the aggregation of different skin-friction contributions still presents challenges due to the significant in-homogeneity in the flow. In this study, a new formulation of the identity derived from the convective form of the governing equations is proposed, considering wall-tangential convection and pressure gradient together. This formulation allows for the identification of different regimes and provides a more effective description of control effects.
INTERNATIONAL JOURNAL OF HEAT AND FLUID FLOW
(2023)
Article
Mechanics
Yonghong Zhong, Kai Fukami, Byungjin An, Kunihiko Taira
Summary: We developed machine learning methods to reconstruct unsteady vortical flow fields from limited sensor measurements, using only a small amount of training data. The machine learning models accurately reconstructed aerodynamic force coefficients, pressure distributions, and vorticity fields for various cases.
THEORETICAL AND COMPUTATIONAL FLUID DYNAMICS
(2023)
Article
Mechanics
Vishal Anatharaman, Jason Feldkamp, Kai Fukami, Kunihiko Taira
Summary: We investigate the compression of spatial and temporal features in fluid flow data using multimedia compression techniques. The effectiveness of spatial compression techniques (including JPEG and JP2) and spatiotemporal video compression techniques (namely H.264, H.265, and AV1) in minimizing compression artifacts and preserving underlying flow physics are examined for different flow scenarios. These compression techniques achieve significant data compression while maintaining dominant flow features with minimal error. AV1 and H.265 compressions demonstrate superior performance across various canonical flow regimes, outperforming traditional techniques like proper orthogonal decomposition in some cases. These image and video compression algorithms are flexible, scalable, and widely applicable in fluid dynamics for data storage and transfer.
THEORETICAL AND COMPUTATIONAL FLUID DYNAMICS
(2023)
Article
Mechanics
Kai Fukami, Koji Fukagata, Kunihiko Taira
Summary: This paper surveys machine-learning-based super-resolution reconstruction for vortical flows. Super resolution aims to find the high-resolution flowfields from low-resolution data and is generally an approach used in image reconstruction. In addition to surveying a variety of recent super-resolution applications, we provide case studies of super-resolution analysis for an example of two-dimensional decaying isotropic turbulence. We demonstrate that physics-inspired model designs enable successful reconstruction of vortical flows from spatially limited measurements. We also discuss the challenges and outlooks of machine-learning-based superresolution analysis for fluid flow applications. The insights gained from this study can be leveraged for superresolution analysis of numerical and experimental flow data.
THEORETICAL AND COMPUTATIONAL FLUID DYNAMICS
(2023)
Article
Engineering, Aerospace
Dashuai Chen, Frieder Kaiser, Jiacheng Hu, David E. Rival, Kai Fukami, Kunihiko Taira
Summary: This study explores the feasibility of using multilayer perceptron (MLP) for estimating aerodynamic loads in complex gusty environments. The results show that the MLP model is able to accurately estimate the relationship between surface pressure and aerodynamic loads, and reveal the importance of sensors located near the leading edge and nose of the wing.
Article
Multidisciplinary Sciences
Kai Fukami, Kunihiko Taira
Summary: As extreme weather conditions become more frequent, it is crucial for small air vehicles to achieve stable flight in the presence of atmospheric disturbances. However, there is a lack of theoretical understanding of the influence of extreme vortical gusts on wings. In this study, machine learning is used to reveal a low-dimensional manifold that captures the extreme aerodynamics of gust-airfoil interactions, enabling real-time reconstruction, modeling, and control of unsteady gusty flows. These findings provide support for the stable flight of next-generation small air vehicles in adverse weather conditions.
NATURE COMMUNICATIONS
(2023)
Article
Engineering, Aerospace
Surabhi Singh, Lawrence Ukeiley, Yang Zhang, Louis Cattafesta, Kunihiko Taira
Summary: This study investigates passive flow control for a rectangular cavity in supersonic flow using a spanwise array of leading-edge tabs. The tabs are shown to effectively reduce pressure fluctuations on the cavity surfaces, with evidence of changes in flow characteristics and the presence of counter-rotating streamwise vortices. Insights provided by the analysis demonstrate that the control mechanisms of the passive tabs are analogous to steady leading-edge blowing strategies.
JOURNAL OF AIRCRAFT
(2022)
Proceedings Paper
Mechanics
Ross Richardson, Brian Eckert, Yang Zhang, Louis N. Cattafesta, Adam Edstrand, Yiyang Sun, Peter Schmid, Kunihiko Taira
Summary: Inspired by a parabolized stability analysis, this study targets higher-order wake instability modes through the actuation of a tensioned string, and the results show increased turbulence intensity in the wake and a reduction in the time-averaged streamwise vorticity downstream with minimal loss of lift, providing evidence of the possibility of exciting instability modes for subsequent attenuation of the vortex farther downstream.
IUTAM LAMINAR-TURBULENT TRANSITION
(2022)