POD-spectral decomposition for fluid flow analysis and model reduction

Cammilleri, Ada; Guéniat, Florimond; Carlier, Johan; Pastur, Luc; Mémin, Étienne; Lusseyran, François; Artana, Guillermo

doi:10.1007/s00162-013-0293-2

Cited by 37 publications

(18 citation statements)

References 32 publications

(58 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The detail and generalities about the methodology are not the subject here, but the reader may find helpful information, for instance, inHolmes, Lumley & Berkooz (1996) andCordier & Bergmann (2003). For examples of POD applications and developments in fluid mechanics, one may notably refer to the recent works ofIungo & Lombardi (2011),Kitsios et al (2011), Buchmann, Atkinson & Soria (2013,Cammilleri et al (2013) and…”

mentioning

confidence: 99%

On the modulating effect of three-dimensional instabilities in open cavity flows

et al. 2014

Self Cite

View full text Add to dashboard Cite

Open cavity flows are known to select and enhance locked-on modes or tones. High-energy self-sustained oscillations arise within the shear layer, impinging onto the trailing edge of the cavity. These self-sustained oscillations are subject to amplitude modulations (AMs) at multiple low frequencies. However, only a few studies have addressed the identification of the lowest modulating frequencies. The present work brings to light salient AMs of the shear layer waves and identifies their source as three-dimensional dynamics existing inside the cavity. Indeed, the recirculating inner flow gives rise to centrifugal instabilities, which entail broad-band frequencies down two orders of magnitude lower than those of the self-sustained oscillations. Using timeresolved PIV (TRPIV) in two planes, the nonlinearly saturated dynamics is analysed in both space and time by means of proper orthogonal decomposition, global Fourier decomposition and Hilbert-Huang transforms. The inner flow can be decomposed as three-dimensional waves carried by the main recirculation. Bicoherence distributions are computed to highlight the nonlinear interactions between these spanwise-travelling waves inside the cavity and the locked-on modes. The modulated envelope of the shear layer oscillations is extracted and investigated with regards to the inner-flow dynamics. Strong cross-correlations, in time rather than in space, reveal a global coupling mechanism, possibly related to the beating of the spanwise-travelling waves.

show abstract

mentioning

confidence: 99%

On the modulating effect of three-dimensional instabilities in open cavity flows

et al. 2014

Self Cite

View full text Add to dashboard Cite

show abstract

“…The amplitude of the modal coefficients decreases as the mode number increases (e.g., the amplitude of the first modal series is higher than the second one) as anticipated. This is further observed by computing the standard deviation of each coefficient time series 11 . where, stands for the amplitude, k for wave number and for angular velocity 12 .…”

Section: Proper Orthogonal Decompositionmentioning

confidence: 99%

Analysis of Turbine Wake Characteristics using Proper Orthogonal Decomposition (POD) and Triple Decomposition Methods

Premaratne

Tian

2016

46th AIAA Fluid Dynamics Conference

View full text Add to dashboard Cite

We examine the flow characteristics behind a commonly-used three-bladed horizontal-axis wind turbine via experiments in a large-scale wind tunnel with a scaled model placed in a typical Atmospheric Boundary Layer (ABL) wind under neutral stability conditions. A high-resolution digital particle image velocimetry (PIV) system was used to study the detailed flow field. Besides analyzing average statistics of the flow quantities such as mean velocity, Reynolds stress, and vorticity distributions in the wake, ''phase-locked'' PIV measurements also elucidated further details of the wake vortex structures for a frozen position of the blades. Proper Orthogonal Decomposition (POD) method was employed to identify the high energy modes that dominate the turbulent kinetic energy distributions in the turbine wakes. Triple Decomposition (TD) was used to elucidate the underlying physics of the intensive turbulent mixing process in the wake flow, promoting the downward transport of high-speed flow. We managed to identify the principal coherent structures in the flow and derived the physics behind vortex break-down and the propagation. Nomenclature= Modal Coefficients for POD a = Amplitude of Kelvin Waves = Eigenvector matrix ABL = Atmospheric boundary layer C P = Power coefficient C T = Thrust coefficient D = Diameter of the rotor H = Hub height HAWT = Horizontal axis wind turbine k = Wave number L = Length from the tower to measurement location P = Power output of the wind turbine PIV = Particle image velocimetry = Reynolds shear stress TKE = Turbulent kinetic energy TSR = Tip speed ratio U = Velocity Magnitude/Snapshot Matrix u,v = Axial and vertical velocity components U H = Inflow velocity at hub height ω = Vorticity Z,Zr = Vertical axis location/Reference location = Power law exponent ∅ = POD modes Γ = Circulation Strength = Stream function

show abstract

“…Another variant uses the combination of Proper Orthogonal Decomposition (POD) and DMD as described in [30]. First, the POD extracts spatial coherent structures (topos), which are ranked by their associated fluctuating energy content.…”

Section: Modeling Robot Dynamics Using Dmdmentioning

confidence: 99%

Estimation of perturbations in robotic behavior using dynamic mode decomposition

et al. 2015

View full text Add to dashboard Cite

Physical human-robot interaction tasks require robots that can detect and react to external perturbations caused by the human partner. In this contribution, we present a machine learning approach for detecting, estimating, and compensating for such external perturbations using only input from standard sensors. This machine learning approach makes use of Dynamic Mode Decomposition (DMD), a data processing technique developed in the field of fluid dynamics, which is applied to robotics for the first time. DMD is able to isolate the dynamics of a nonlinear system and is therefore well suited for separating noise from regular oscillations in sensor readings during cyclic robot movements. In a training phase, a DMD model for behavior-specific parameter configurations is learned. During task execution, the robot must estimate the external forces exerted by a human interaction partner. We compare the DMD-based approach to other interpolation schemes. A variant, sparsity promoting DMD, is particularly well suited for high-noise sensors. Results of a user study show that our DMD-based machine learning approach can be used to design physical human-robot interaction techniques that not only result in robust robot behavior but also enjoy a high usability.

show abstract

POD-spectral decomposition for fluid flow analysis and model reduction

Cited by 37 publications

References 32 publications

On the modulating effect of three-dimensional instabilities in open cavity flows

On the modulating effect of three-dimensional instabilities in open cavity flows

Analysis of Turbine Wake Characteristics using Proper Orthogonal Decomposition (POD) and Triple Decomposition Methods

Estimation of perturbations in robotic behavior using dynamic mode decomposition

Contact Info

Product

Resources

About