Data-driven feature identification and sparse representation of turbulent flows

Identifying coherent structures in fluid flows is of great importance for reduced order modelling and flow control. However, extracting such structures from experimental or numerical data obtained from a turbulent flow can be challenging. A number of modal decomposition algorithms have been proposed in recent years which decompose time-resolved snapshots of data into spatial modes, each associated with a single frequency and growth-rate. Most prominently among them is dynamic mode decomposition (DMD). However, DMD-like algorithms create an arbitrary number of modes. It is common practice to then choose a smaller subset of these modes, for the purpose of model reduction and analysis, based on some measure of significance. In this work, we present a method of post-processing DMD modes for extracting a small number of dynamically relevant modes. We achieve this through an iterative approach based on the graph-theoretic notion of maximal cliques to identify clusters of modes and representing each cluster with a single representative mode.     

Energy balance of reduced order models for unsteady flows using sparse coding

Galerkin projection is a commonly used reduced order modeling approach; however, stability and accuracy of the resulting models are open issues for unsteady flow fields. Balance between production and dissipation of energy is crucial for stability. Moreover, the rates of energy production and dissipation are functions of large- and small-scale information captured in the chosen modes. Due to the highly nonlinear nature of the Navier–Stokes equations, the process of choosing an ‘appropriate’ set of modes from the simulation or experimental data is non-trivial. Recent work indicates that modal decompositions computed using a sparse coding approach yield multi-scale modes that provide improved low-order models compared to the commonly used proper orthogonal decomposition. This study seeks to use energy components analysis to develop a deeper understanding of the improved model performance with sparse modes. In addition, a greedy search-based sparse coding algorithm is developed for basis extraction. The analysis is performed on two canonical problems of incompressible flows inside a lid-driven cavity and past a stationary cylinder. Results indicate that there is a direct link between the presence of multi-scale features in the reduced set of modes, balance between production and dissipation of energy, and reduced order model performance.     

An improved algorithm for the shallow water equations model reduction: Dynamic Mode Decomposition vs POD

We propose an improved framework for dynamic mode decomposition (DMD) of 2‐D flows for problems originating from meteorology when a large time step acts like a filter in obtaining the significant Koopman modes, therefore, the classic DMD method is not effective. This study is motivated by the need to further clarify the connection between Koopman modes and proper orthogonal decomposition (POD) dynamic modes. We apply DMD and POD to derive reduced order models (ROM) of the shallow water equations. Key innovations for the DMD‐based ROM introduced in this paper are the use of the Moore–Penrose pseudoinverse in the DMD computation that produced an accurate result and a novel selection method for the DMD modes and associated amplitudes and Ritz values. A quantitative comparison of the spatial modes computed from the two decompositions is performed, and a rigorous error analysis for the ROM models obtained is presented. Copyright © 2015 John Wiley & Sons, Ltd.     

On the use of sensitivity analysis in model reduction to predict flows for varying inflow conditions

The proper orthogonal decomposition (POD)‐based model reduction method is more and more successfully used in fluid flows. However, the main drawback of this methodology rests in the robustness of these reduced order models (ROMs) beyond the reference at which POD modes have been derived. Any variation in the flow or shape parameters within the ROM fails to predict the correct dynamics of the flow field. To broaden the spectrum of these models, the POD modes should have the global characteristics of the flow field over which the predictions are required. Mixing of snapshots with varying parameters is one way to improve the global nature of the modes but is computationally demanding because it requires full‐order solutions for a number of parameter values in order to assemble atextitrich enough database on which to perform POD. Instead, we have used sensitivity analysis (SA) to include the flow and shape parameters influence during the basis selection process to develop more robust ROMs for varying viscosity (Reynolds number), changing orientation and shape definition of bodies. This study aims at extending these ideas to inflow conditions to demonstrate the effectiveness of the proposed approach in capturing the effect of varying inflow on the dynamics of the flow over an elliptic cylinder. Numerical experiments show that the newly derived models allow for a more accurate representation of the flows when exploring the parameter space. Copyright © 2011 John Wiley & Sons, Ltd.     

Direct measurement of the stream-function in a quasi-two-dimensional liquid metal flow

 Flows of electrically conducting fluids become quasi-two-dimensional (2D) under the influence of a homogeneous magnetic field. In this regime the electrical potential induced in the flow is directly proportional to the 2D stream function. We utilize this property in order to perform simultaneous multi-channel measurements of the stream function in an electromagnetically driven annular liquid metal flow in the weakly nonlinear regime. Spatio-temporal maps of the stream functions of a quasi-periodic unstable flow are observed and used as a basis for a systematic decomposition of the flow into spatial modes belonging to different frequencies. The approach can be generalized to flows with complex time-dependence. Received: 3 March 1997/Accepted: 27 September 1997     

A finite element formulation for global linear stability analysis of a nominally two‐dimensional base flow

A stabilized finite element method, to carry out the linear stability analysis of a two‐dimensional base flow to three‐dimensional perturbations that are periodic along span, is presented. The resulting equations for the time evolution of the disturbance requires a solution to the generalized eigenvalue problem. The analysis is global in nature and is also applicable to non‐parallel flows. Equal‐order‐interpolation functions for velocity and pressure are utilized. Stabilization terms are added to the Galerkin formulation to admit the use of equal‐order‐interpolation functions and to eliminate node‐to‐node oscillations that might arise in advection‐dominated flows. The proposed formulation is tested on two flow problems. First, the mode transitions in the circular Couette flow are investigated. Two scenarios are considered. In the first one, the outer cylinder is at rest, while the inner one spins. Two linearly unstable modes are identified. The primary mode is real and represents the axisymmetric Taylor vortices. The second mode is complex and consists of spiral vortices. For the counter‐rotating cylinders, the primary transition is via the appearance of spiral vortices. Excellent agreement with results from earlier studies is observed. The formulation is also utilized to investigate the parallel and oblique modes of vortex shedding past a cylinder for the Re = 100 flow. It is found that the flow is associated with a large number of unstable oblique shedding modes. The parallel mode of vortex shedding is a special case of this family of modes and is associated with the largest growth rate. Copyright © 2014 John Wiley & Sons, Ltd.     

On Low-Dimensional Modeling of Channel Turbulence

We investigate a sequence of low-dimensional models of turbulent channel flows. These models are based on the extraction of the Karhunen–Loève (KL) eigenfunctions from a large-scale simulation in a wide channel with R _*=180 (based on the friction velocity). KL eigenfunctions provide an optimal coordinate system in which to represent the dynamics of the turbulent flow. The hierarchy of KL modes identifies the most energetic independent phenomena in the system. A series of Galerkin projections is then used to derive a dynamical approximation to flows. In order to capture essential aspects of the flow in a low-dimensional system, a careful selection of modes is carried out. The resulting models satisfy momentum and energy conservation as well as yielding the amount of viscous dissipation found in the exact system. Their dynamics includes modes which characterize the flux, rolls, and propagating waves. Unlike previous treatments the instantaneous streamwise flow is time dependent and represented by KL flux modes. The rolls correspond to the streaks observed in experiments in the viscous sublayer. Propagating waves which first appear in the model flow at low Reynolds number (R _*∼ 60) persist through the chaotic regime that appears as the Reynolds number is increased. Statistical measures of the chaotic flows which have been generated by the models compare favorably with those found in full-scale simulations. Received 13 July 1998 and accepted 8 January 1999     

无壁面参数低雷诺数非线性涡黏性模式研究

建立了一个低雷诺数的非线性涡黏性湍流模式,该模式的一个显著特性是它不包含壁面参数（如y^ ,n等）,因而特别适用于复杂几何流场的计算,本模式在几种包括回流、分离、激波等典型流动中进行了验证,结果令人满意。     

A reduced‐order approach for optimal control of fluids using proper orthogonal decomposition

In this article, a reduced‐order modeling approach, suitable for active control of fluid dynamical systems, based on proper orthogonal decomposition (POD) is presented. The rationale behind the reduced‐order modeling is that numerical simulation of Navier–Stokes equations is still too costly for the purpose of optimization and control of unsteady flows. The possibility of obtaining reduced‐order models that reduce the computational complexity associated with the Navier–Stokes equations is examined while capturing the essential dynamics by using the POD. The POD allows the extraction of a reduced set of basis functions, perhaps just a few, from a computational or experimental database through an eigenvalue analysis. The solution is then obtained as a linear combination of this reduced set of basis functions by means of Galerkin projection. This makes it attractive for optimal control and estimation of systems governed by partial differential equations (PDEs). It is used here in active control of fluid flows governed by the Navier–Stokes equations. In particular, flow over a backward‐facing step is considered. Reduced‐order models/low‐dimensional dynamical models for this system are obtained using POD basis functions (global) from the finite element discretizations of the Navier–Stokes equations. Their effectiveness in flow control applications is shown on a recirculation control problem using blowing on the channel boundary. Implementational issues are discussed and numerical experiments are presented. Copyright © 2000 John Wiley & Sons, Ltd.     

Spatially developing secondary instabilities in compressible swept airfoil boundary layers

Two-dimensional eigenvalue analysis is used on a massive scale to study the spatial instabilities of compressible shear flows with two inhomogeneous directions. The main focus of the study is crossflow dominated swept-wing boundary layers although the methodology can also be applied to study other types of flows, such as the attachment-line flow. Certain unique aspects of formulating a spatial, two-dimensional eigenvalue problem for the secondary instability of finite amplitude crossflow vortices are discussed, namely, fixing the spatial growth direction unambiguously through a non-orthogonal formulation of the linearized disturbance equations. A primary test case used for parameter study corresponds to Numerical results are presented for the low-speed, NLF-0415(b) airfoil configuration as tested in the ASU Unsteady Wind Tunnel, wherein a spanwise periodic array of roughness elements was placed near the leading edge in order to excite stationary crossflow modes with a specified fundamental wavelength. The two classes of flow conditions selected for this analysis include those for which the roughness array spacing corresponds to either the naturally dominant crossflow wavelength, or a subcritical wavelength that serves to reduce the growth of the naturally excited dominant crossflow modes. Numerical predictions are compared with the measured database, both as indirect validation for the spatial instability analysis and to provide a basis for comparison with a higher Reynolds number, supersonic swept-wing configuration. Application of the eigenvalue analysis to the supersonic configuration reveals that a broad spectrum of stationary crossflow modes can sustain sufficiently strong secondary instabilities as to potentially cause transition over this configuration. In particular, the control mode itself, if initiated with too large an amplitude, may lead to an earlier transition.     

Flow of an electrorheological fluid between eccentric rotating cylinders

Electrorheological fluids have numerous potential applications in vibration dampers, brakes, valves, clutches, exercise equipment, etc. The flows in such applications are complex three-dimensional flows. Most models that have been developed to describe the flows of electrorheological fluids are one-dimensional models. Here, we discuss the behavior of two fully three-dimensional models for electrorheological fluids. The models are such that they reduce, in the case of simple shear flows with the intensity of the electric field perpendicular to the streamlines, to the same constitutive relation, but they would not be identical in more complicated three-dimensional settings. In order to show the difference between the two models, we study the flow of these fluids between eccentrically placed rotating cylinders kept at different potentials, in the setting that corresponds to technologically relevant problem of flow of electrorheological fluid in journal bearing. Even though the two models have quite a different constitutive structure, due to the assumed forms for the velocity and pressure fields, the models lead to the same velocity field but to different pressure fields. This finding illustrates the need for considering the flows of fluids described by three-dimensional constitutive models in complex geometries, and not restricting ourselves to flows of fluids described by one-dimensional models or simple shear flows of fluids characterized by three-dimensional models.     

Strongly coupled simulation of fluid–structure interaction in a Francis hydroturbine

This work simulates a complex fluid flow in fluid–structure interaction (FSI). The flow under consideration is governed by Navier–Stokes equations for incompressible viscous fluids and modeled with the finite volume method. Large eddy simulation is used to simulate the unsteady turbulent flow. The structure is represented by a finite element formulation. The present work introduces a strongly coupled partitioned approach that is applied to complex flow in fluid machinery. In this approach, the fluid and structure equations are solved separately using different solvers, but are implicitly coupled into one single module based on sensitivity analysis of the important displacement and stress modes. The applied modes and their responses are used to build up a reduced‐order model. The proposed model is used to predict the unsteady flow fields of a 3D complete passage, involving in stay, guide vanes, and runner blades, for a Francis hydro turbine and FSI is considered. The computational results show that a fairly good convergence solution is achieved by using the reduced‐order model that is based on only a few displacement and stress modes, which largely reduces the computational cost, compared with traditional approaches. At the same time, a comparison of the numerical results of the model with available experimental data validates the methodology and assesses its accuracy. Copyright © 2008 John Wiley & Sons, Ltd.     

Flow Structures Associated with Turbulent Mixing Noise and Screech Tones in Axisymmetric Jets

A jet from an axisymmetric convergent nozzle is studied at ideal and underexpanded conditions using velocity and acoustic data. Two particle imaging velocimetry setups, a 10 kHz system and a multi-camera configuration, were used to capture near-field velocities while simultaneously sampled with far-field microphones. Proper orthogonal decomposition is performed on the velocity data to extract modes representative of physical processes in the flow. The decomposed velocity fields are then correlated with acoustic data to identify modes related to specific noise spectra. Specifically, four modes are associated with noise production in the sonic plume. Selective flow-field reconstruction is carried out, revealing interesting dynamics associated with loud flow states. In the supersonic case, screech-containing and turbulent mixing modes are isolated. The spatial modes of each data set are then compared for similarities in structures.     

Spatial Evolution of Nonstationary Perturbations in Hamel Flow

The propagation of small perturbations in longitudinally inhomogeneous flows is investigated. The evolution of the perturbations is studied with reference to the radial flow of a viscous incompressible fluid between plane nonparallel walls, the simplest inhomogeneous flow. Using a generalized method of variation of constants, the corresponding boundary-value problem is reduced to an infinite-dimensional evolutionary system of ordinary differential equations for the complex amplitudes of the eigensolutions of a locally homogeneous problem. Physically, the method can be interpreted as a representation of the perturbation evolution process via two concomitant processes: the independent amplification (attenuation) of normal modes of the locally homogeneous problem and the rescattering of these modes into each other on local inhomogeneities of the base flow. The calculations show that reduced versions of the method are promising for describing the linear stage of laminar-turbulent transition in a boundary layer.     

Modal decomposition‐based global stability analysis for reduced order modeling of 2D and 3D wake flows

The method for computation of stability modes for two‐ and three‐dimensional flows is presented. The method is based on the dynamic mode decomposition of the data resulting from DNS of the flow in the regime close to stable flow (fixed‐point dynamics, small perturbations about steady flow). The proposed approach is demonstrated on the wake flows past a 2D, circular cylinder, and a sphere. The resulting modes resemble the eigenmodes computed conventionally from global stability analysis and are used in model order reduction of the flow. The designed low‐dimensional Galerkin model uses continuous mode interpolation between dynamic mode decomposition mode bases and reproduces the dynamics of Navier–Stokes equations. Copyright © 2015 John Wiley & Sons, Ltd.     

Low-Dimensional Modelling of Turbulence Using the Proper Orthogonal Decomposition: A Tutorial

The proper orthogonal decomposition identifies basis functions or modes which optimally capture the average energy content from numerical or experimental data. By projecting the Navier–Stokes equations onto these modes and truncating, one can obtain low-dimensional ordinary differential equation models for fluid flows. In this paper we present a tutorial on the construction of such models. In addition to providing a general overview of the procedure, we describe two different ways to numerically calculate the modes, show how symmetry considerations can be exploited to simplify and understand them, comment on how parameter variations are captured naturally in such models, and describe a generalization of the procedure involving projection onto uncoupled modes that allow streamwise and cross-stream components to evolve independently. We illustrate for the example of plane Couette flow in a minimal flow unit – a domain whose spanwise and streamwise extent is just sufficient to maintain turbulence.     

A Bayesian Calibration–Prediction Method for Reducing Model-Form Uncertainties with Application in RANS Simulations

Model-form uncertainties in complex mechanics systems are a major obstacle for predictive simulations. Reducing these uncertainties is critical for stake-holders to make risk-informed decisions based on numerical simulations. For example, Reynolds-Averaged Navier-Stokes (RANS) simulations are increasingly used in the design, analysis, and safety assessment of mission-critical systems involving turbulent flows. However, for many practical flows the RANS predictions have large model-form uncertainties originating from the uncertainty in the modeled Reynolds stresses. Recently, a physics-informed Bayesian framework has been proposed to quantify and reduce model-form uncertainties in RANS simulations for flows by utilizing sparse observation data. However, in the design stage of engineering systems, when the system or device has not been built yet, measurement data are usually not available. In the present work we extend the original framework to scenarios where there are no available data on the flow to be predicted. In the proposed method, we first calibrate the model discrepancy on a related flow with available data, leading to a statistical model for the uncertainty distribution of the Reynolds stress discrepancy. The obtained distribution is then sampled to correct the RANS-modeled Reynolds stresses for the flow to be predicted. The extended framework is a Bayesian calibration–prediction method for reducing model-form uncertainties. The merits of the proposed method are demonstrated on two flows that are challenging to standard RANS models. By not requiring observation data on the flow to be predicted, the present calibration–prediction method will gain wider acceptance in practical engineering design and analysis compared to the original framework. While RANS modeling is chosen to demonstrate the merits of the proposed framework, the methodology is generally applicable to other complex mechanics models involving solids, fluids flows, or the coupling between the two (e.g., mechanics models for the cardiovascular systems), where model-form uncertainties are present in the constitutive relations.     

Bi-linear reduced-order models of structures with friction intermittent contacts

The dynamic analysis of structures with localized nonlinearities, such as intermittent contacts of cracked structures, is a computationally demanding task because of the large size of the models involved. Thus, high-resolution finite element models are often reduced using a variety of specialized techniques which exploit spatial coherences in the dynamics. In addition, when a steady-state forced response analysis is performed, direct time integration can be replaced with multi-harmonic balance methods. Recently, a technique based on bi-linear normal modes has been successfully applied to piecewise-linear oscillators. The key idea of that approach is to represent the spatial coherences in the system dynamics with two sets of normal modes with special boundary conditions, referred to as bi-linear modes. In this paper, the bi-linear modal representation is extended to the case of intermittent contacts with friction. Furthermore, a novel reduced order modeling method is developed for the 0th order harmonic used in multi-harmonic balance methods. The forced response of a cracked structure is used to demonstrate the proposed methods.     

Transient growth mechanisms of low Reynolds number flow over a low-pressure turbine blade

A direct transient growth analysis for three-dimensional perturbations to flow past a periodic array of T-106/300 low-pressure turbine fan blades is presented. The methodology is based on a singular value decomposition of the flow evolution operator, linearised about a steady or periodic base flow. This analysis yields the optimal growth modes. Previous work on global mode stability analysis of this flow geometry showed the flow is asymptotically stable, indicating a non-modal explanation of transition may be more appropriate. The present work extends previous investigations into the transient growth around a steady base flow, to higher Reynolds numbers and periodic base flows. It is found that the notable transient growth of the optimal modes suggests a plausible route to transition in comparison to modal growth for this configuration. The spatial extent and localisation of the optimal modes is examined and possible physical triggering mechanisms are discussed. It is found that for longer times and longer spanwise wavelengths, a separation in the shear layer excites the wake mode. For shorter times and spanwise wavelengths, smaller growth associated with excitation of the near wake are observed.     

Introduction of turbulent model in a mixed finite volume/finite element method

The aim of this work is to present a new numerical method to compute turbulent flows in complex configurations. With this in view, a k-? model with wall functions has been introduced in a mixed finite volume/finite element method. The numerical method has been developed to deal with compressible flows but is also able to compute nearly incompressible flows. The physical model and the numerical method are first described, then validation results for an incompressible flow over a backward-facing step and for a supersonic flow over a compression ramp are presented. Comparisons are performed with experimental data and with other numerical results. These simulations show the ability of the present method to predict turbulent flows, and this method will be applied to simulate complex industrial flows (flow inside the combustion chamber of gas turbine engines). The main goal of this paper is not to test turbulence models, but to show that this numerical method is a solid base to introduce more sophisticated turbulence model.