Article
Computer Science, Interdisciplinary Applications
Quincy A. Huhn, Mauricio E. Tano, Jean C. Ragusa, Youngsoo Choi
Summary: Dynamic Mode Decomposition (DMD) is a model-order reduction technique that extracts spatial modes of fixed temporal frequencies from numerical or experimental data. This paper presents two novel approaches to parametric DMD: one based on interpolation of the reduced-order DMD eigen-pair and the other based on interpolation of the reduced DMD (Koopman) operator. Numerical results are provided for diffusion-dominated nonlinear dynamical problems, including a multiphysics radiative transfer example. The three parametric DMD approaches are compared.
JOURNAL OF COMPUTATIONAL PHYSICS
(2023)
Article
Chemistry, Analytical
Zhangfeng Zhao, Gaohong Liu, Yueliang Wang, Jiyu Peng, Xin Qiao, Jiang Zhong
Summary: This study proposes a denoising method using variational mode decomposition combined with a wavelet threshold to improve the accuracy of tea flow measurement. Compared with other methods, the VMD-WT method demonstrates superior denoising performance and provides an effective method for dynamic and accurate measurement of tea flow in tea processing.
Article
Engineering, Marine
Yadong Han, Ming Liu, Lei Tan
Summary: This study investigates the characteristics of the cavitating flow in a Venturi nozzle through experimental measurement and simulation analysis. POD and DMD methods are applied to study the three-dimensional structures of cavitation and flow fields, revealing the cavitation-velocity interaction based on the mode decomposition of vapor volume fraction and streamwise velocity.
Article
Thermodynamics
Fan Yang, Tao Wu, Hui Jiang, Jinyang Jiang, Hanxue Hao, Lianqiang Zhang
Summary: This study proposed a new HST prediction method based on dynamic mode decomposition (DMD) which can accurately predict transformer's hot spot temperature (HST) and winding temperature field distribution in a few seconds with clear physical meaning.
CASE STUDIES IN THERMAL ENGINEERING
(2022)
Article
Computer Science, Software Engineering
Duong B. Nguyen, Panruo Wu, Rodolfo Ostilla Monico, Guoning Chen
Summary: Large-scale structures in shear flows play a significant role in understanding physical phenomena and modeling complex turbulence flows. To address the limitations of conventional methods, we propose the use of Multi-Resolution Dynamic Mode Decomposition (mrDMD) to extract large-scale structures in shear flows. Our method utilizes slow motion DMD modes to capture the dynamics of these structures and provides an efficient way to visualize them. We also provide a GPU-based implementation to speed up the computation of mrDMD.
IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS
(2023)
Article
Mathematics
Keren Li, Sergey Utyuzhnikov
Summary: In this paper, a low-rank tensor decomposition-based implementation of HODMD is proposed to reduce computational complexity and provide a more efficient dynamic mode decomposition method. The efficiency of this method is demonstrated with examples, including load forecasting of a power system, showing comparisons with respect to computing time and accuracy.
Article
Energy & Fuels
Yanzhao Wu, Di Zhu, Ran Tao, Ruofu Xiao, Weichao Liu
Summary: This article discusses the problem of cavitation in the design, manufacture, and operation of pump-turbines and the limitations of traditional research methods. The dynamic mode decomposition (DMD) method is used to study the cavitation of a model pump-turbine and accurately extract the characteristics of the runner under cavitation conditions. The development of cavitation significantly affects the reconstruction of low order modes. The article provides a new research idea for cavitation research of pump-turbines.
JOURNAL OF ENERGY STORAGE
(2022)
Article
Mechanics
Chi Zhang, Chao Tao, Heng Song, Xiao Han, Lei Li, Xunchen Liu, Fei Qi
Summary: This study investigates the effect of central vortex core (CVC) on the flow and flame in a centrally staged swirl spray combustor. It is found that the CVC structure exists in an extended strip area of strong vorticity near the centerline of the combustor. The motion of CVC is mainly swing, followed by precessing.
Article
Engineering, Mechanical
Akira Saito, Masato Tanaka
Summary: This paper presents a data-driven model order reduction method for piecewise-linear systems based on dynamic mode decomposition (DMD). The method is applied to two representative examples of piecewise linear systems. By extracting dynamic modes and projecting them onto a subspace, the proposed method reduces the degrees of freedom of the system. The reduced order models constructed by this method produce accurate forced response results.
NONLINEAR DYNAMICS
(2023)
Article
Computer Science, Interdisciplinary Applications
Zhen Gao, Yifan Lin, Xiang Sun, Xueying Zeng
Summary: Accurately constructing a reduced order model (ROM) of nonlinear parameterized partial differential equations (PDEs) has always been a challenging problem. In this paper, a new ROM method based on KNN-DMD is proposed, which demonstrates good applicability, efficiency, and predictive ability for parameterized PDEs.
JOURNAL OF COMPUTATIONAL PHYSICS
(2022)
Article
Mechanics
H. K. Jang, C. E. Ozdemir, J-H Liang, M. Tyagi
Summary: This study applies Dynamic Mode Decomposition (DMD) to investigate the oscillatory flow around a vertical wall-mounted cylinder, demonstrating its effectiveness in capturing dynamic and nonlinear flow features, and providing a basis for data-driven models relevant to coastal engineering applications.
Article
Engineering, Mechanical
Adrian Corrochano, Giuseppe D'Alessio, Alessandro Parente, Soledad Le Clainche
Summary: This work presents a new application of higher order dynamic mode decomposition (HODMD) combined with preprocessing techniques such as principal component analysis (PCA) for the analysis of reactive flows. The new methodology proves to be effective in identifying the main patterns and developing reduced order models. The algorithm is also coupled with feature selection step carried out via PCA and varimax rotation, showing outstanding capabilities in compressing original databases.
INTERNATIONAL JOURNAL OF MECHANICAL SCIENCES
(2023)
Article
Engineering, Aerospace
Van Duc Nguyen, Viet Dung Duong, Minh Hoang Trinh, Hoang Quan Nguyen, Dang Thai Son Nguyen
Summary: This paper presents low order modeling, including iterative Brinkman penalized vortex method (IBVM) and data-driven dynamic mode decomposition (DMD), for studying dynamic stall of a symmetric airfoil. Data from IBVM is extracted and used for flow field reconstruction using dominant modes in DMD, representing extracted flow features. The dominant modes, consisting of the primary mode, its harmonics, and the mean mode, are found to be crucial for wake duplication of the airfoil at fixed angles of attack. The DMD results show good agreement with IBVM and experimental data for nearfield and farfield vorticity contours, as well as the lift coefficient.
INTERNATIONAL JOURNAL OF MICRO AIR VEHICLES
(2023)
Article
Chemistry, Analytical
Zhiwen Lu, Feng Li, Shancheng Cao, Rui Yuan, Yong Lv
Summary: A novel crack localization method for operating rotors is proposed, which can accurately and robustly achieve multi-crack localization without baseline information, and eliminate common interferences.
Article
Energy & Fuels
L. X. Wang, J. H. Zheng, Z. G. Li, Z. X. Jing, Q. H. Wu
Summary: This paper proposes an order reduction method (ORM) to handle the complex dynamics of heterogeneous integrated energy systems (HIES). The ORM combines model partitioning and modal synthesis techniques to effectively simulate high-order dynamics, with advantages in both computation accuracy and time.
Article
Mechanics
Yinzhu Wang, Esteban Ferrer, Jorge Saavedra, Guillermo Paniagua, Eusebio Valero
Summary: This study investigates an optimization approach using blowing to prevent flow separation at low Reynolds numbers, comparing a stability-based method with a traditional pressure loss-related optimization. The stability-based method showed comparable results to the traditional approach, demonstrating its potential usefulness in cases with unclear integral functionals.
Article
Computer Science, Interdisciplinary Applications
Andres M. Rueda-Ramirez, Esteban Ferrer, David A. Kopriva, Gonzalo Rubio, Eusebio Valero
Summary: A static condensation method is presented for improving computational efficiency and enhancing the condition number of linear systems. The statically condensed GL-DGSEM can be applied to various advection-diffusion equations and demonstrates high efficiency and speed-ups when solving the Navier-Stokes equations.
JOURNAL OF COMPUTATIONAL PHYSICS
(2021)
Article
Computer Science, Interdisciplinary Applications
Jiaqing Kou, Saumitra Joshi, Aurelio Hurtado-de-Mendoza, Kunal Puri, Charles Hirsch, Esteban Ferrer
Summary: In this work, the numerical advantages of the high-order Flux Reconstruction (FR) method and the simplicity of the mesh generation of the Immersed Boundary Method (IBM) are combined for steady and unsteady problems over moving geometries using the volume penalization (penalty-IBM) method. The efficiency of the approach in handling moving geometries is demonstrated through various numerical test cases, showcasing the potential of this method in industrial design processes.
JOURNAL OF COMPUTATIONAL PHYSICS
(2022)
Article
Computer Science, Interdisciplinary Applications
Wojciech Laskowski, Gonzalo Rubio, Eusebio Valero, Esteban Ferrer
JOURNAL OF COMPUTATIONAL PHYSICS
(2022)
Article
Computer Science, Interdisciplinary Applications
Jiaqing Kou, Aurelio Hurtado-de-Mendoza, Saumitra Joshi, Soledad Le Clainche, Esteban Ferrer
Summary: This paper presents eigensolution and non-modal analyses for IBM based on volume penalization for the linear advection equation, investigating the influence of penalization parameter on numerical errors and stability. Through semi-discrete and fully-discrete analyses, practical guidelines are provided for selecting penalty parameters, along with a proposal to include a second-order term for improved accuracy and relaxed time step restrictions. Results suggest that careful selection of penalty terms and damping can lead to a more accurate scheme for simulating flow past a cylinder.
JOURNAL OF COMPUTATIONAL PHYSICS
(2022)
Article
Computer Science, Interdisciplinary Applications
Jiaqing Kou, Soledad Le Clainche, Esteban Ferrer
Summary: This study introduces a data-driven method to conduct eigensolution analyses and quantify numerical errors. The new approach, based on Spatio-Temporal Koopman Decomposition, accurately predicts dispersion-dissipation behavior in eigensolution analyses.
JOURNAL OF COMPUTATIONAL PHYSICS
(2022)
Article
Computer Science, Interdisciplinary Applications
Aurelio Hurtado-de-Mendoza, Jiaqing Kou, Saumitra Joshi, Kunal Puri, Charles Hirsch, Esteban Ferrer
Summary: We present a numerical analysis of linear multigrid operators for the high-order Flux Reconstruction method. The non-modal analysis is used to assess the short-term numerical dissipation in the context of 1D and 2D linear convection-diffusion. The effect of several parameters, including the number of coarse-level iterations, the polynomial order, and the combination of h- and p-multigrid, is explored to find the most efficient configurations. V-cycle p-multigrid is shown to be more efficient at higher polynomial orders, and the use of W-cycles and/or hp-multigrid appears to offer additional advantages. Additionally, the influence of high Peclet numbers and high aspect-ratio cells on error dissipation is investigated in the 2D case. Finally, the non-modal dissipation is related to the convergence rate of the multigrid through a series of manufactured solutions.
JOURNAL OF COMPUTATIONAL PHYSICS
(2022)
Article
Computer Science, Interdisciplinary Applications
Fernando Manrique de Lara, Esteban Ferrer
Summary: High order discontinuous Galerkin methods offer accurate solutions with high order polynomials, but at increased cost. This study proposes using Neural Networks to accelerate the method by training the network to correct low order solutions and recover high order accuracy. The results show good accuracy and acceleration, particularly for high polynomial orders.
COMPUTERS & FLUIDS
(2022)
Article
Computer Science, Interdisciplinary Applications
Diego Lodares, Juan Manzanero, Esteban Ferrer, Eusebio Valero
Summary: This article presents an entropy-stable formulation for the compressible Reynolds Averaged Navier-Stokes (RANS) equations and the Spalart-Allmaras closure model. The proposed model satisfies an entropy law and employs a high-order Discontinuous Galerkin (DG) approximation with discrete stability analysis. The schemes demonstrate stability and accuracy for three-dimensional unstructured meshes and different flow cases.
JOURNAL OF COMPUTATIONAL PHYSICS
(2022)
Article
Computer Science, Interdisciplinary Applications
Gerasimos Ntoukas, Juan Manzanero, Gonzalo Rubio, Eusebio Valero, Esteban Ferrer
Summary: This paper presents a novel entropy-stable discontinuous Galerkin approximation for the incompressible Navier-Stokes/Cahn-Hilliard system. The scheme supports p-adaptation and maintains the properties of the original conforming scheme when transitioning to p-non-conforming elements. The authors also introduce a heuristic adaptation criterion and verify the scheme through simulations.
JOURNAL OF COMPUTATIONAL PHYSICS
(2022)
Correction
Computer Science, Interdisciplinary Applications
Diego Lodares, Juan Manzanero, Esteban Ferrer, Eusebio Valero
JOURNAL OF COMPUTATIONAL PHYSICS
(2022)
Article
Computer Science, Interdisciplinary Applications
Jiaqing Kou, Esteban Ferrer
Summary: There has been a recent growing interest in developing efficient immersed boundary methods (IBM) based on Cartesian grids in the context of high-order methods. This work proposes an improvement for high-order solvers based on flux reconstruction by introducing a selective frequency damping (SFD) method to suppress spurious oscillations. Numerical properties are studied through eigensolution analysis and demonstrate the advantages of using the SFD method as an alternative to traditional volume penalization, as well as the beneficial properties of combining both approaches. The new approach is applied to simulate steady and unsteady flow scenarios, showing the improved accuracy provided by the SFD method.
JOURNAL OF COMPUTATIONAL PHYSICS
(2023)
Article
Mechanics
Kheir-Eddine Otmani, Gerasimos Ntoukas, Oscar A. Marino, Esteban Ferrer
Summary: We propose an invariant feature space using principal invariants of strain and rotational rate tensors for detecting viscous-dominated and turbulent regions. The feature space is independent of coordinate frame and allows identification of viscous-dominated rotational region and inviscid irrotational region. Tests on laminar and turbulent flow cases showed that Gaussian mixture clustering effectively identifies these regions without requiring an arbitrary threshold like traditional sensors.
Article
Computer Science, Artificial Intelligence
Jiaqing Kou, Laura Botero-Bolivar, Roman Ballano, Oscar Marino, Leandro de Santana, Eusebio Valero, Esteban Ferrer
Summary: This study presents a framework for optimizing the airfoil shape of wind turbine blades to reduce trailing edge noise. The framework uses Amiet's theory, the TNO-Blake model, and XFOIL simulations to evaluate noise and boundary layer parameters. Particle swarm optimization is employed to find the optimized airfoil configuration, while traditional shape optimization techniques are compared to machine learning methods using a variational autoencoder. The autoencoder-based optimized airfoil reduces overall sound pressure level by 3% (1.75 dBA) and improves aerodynamic properties compared to the baseline NACA0012 airfoil.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Interdisciplinary Applications
Saumitra Joshi, Jiaqing Kou, Aurelio Hurtado de Mendoza, Kunal Puri, Charles Hirsch, Gonzalo Rubio, Esteban Ferrer
Summary: We propose a strategy to estimate the maximum stable time-steps for explicit time-stepping methods for hyperbolic systems in a high-order flux reconstruction framework. The strategy directly incorporates the spatial polynomial-and mesh-discretization in estimating the convective and diffusive length-scales, and is extended to the density-based Navier-Stokes system of equations, taking into account the omnidirectionality of the speed of sound.
COMPUTERS & FLUIDS
(2023)