SEARCH · Search NASA

Results for “Proper orthogonal decomposition”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 55 records · Page 3

All sources NASA NTRS DOE OSTI

DOE OSTI2025

Greedy emulators for nuclear two-body scattering

Applications of reduced basis method emulators are increasing in low-energy nuclear physics because they enable fast and accurate sampling of high-fidelity calculations, enabling robust uncertainty quantification. Here, in this paper, we develop, implement, and test two model-driven emulators based on the (Petrov-)Galerkin projection using the prototypical test case of two-body scattering with the Minnesota potential and a more realistic local chiral potential. The high-fidelity scattering equations are solved with the matrix Numerov method, a reformulation of the popular Numerov recurrence relation for solving special second-order differential equations as a linear system of coupled equations. A novel error estimator based on reduced-space residuals is applied to an active learning approach (a greedy algorithm) to choosing training samples (“snapshots”) for the emulator and contrasted with a proper orthogonal decomposition (POD) approach. Both approaches allow for computationally efficient offline-online decompositions, but the greedy approach requires many fewer snapshot calculations. These developments set the groundwork for emulating scattering observables based on chiral nucleon-nucleon and three-nucleon interactions and optical models, where computational speed-ups are necessary for Bayesian uncertainty quantification. Our emulators and error estimators are widely applicable to linear systems.

Bayesian methods↗

DOE OSTI2024

S-OPT: A Points Selection Algorithm for Hyper-Reduction in Reduced Order Models

While projection-based reduced order models can reduce the dimension of full order solutions, the resulting reduced models may still contain terms that scale with the full order dimension. Hyper-reduction techniques are sampling-based methods that further reduce this computational complexity by approximating such terms with a much smaller dimension. The goal of this work is to introduce the points selection algorithm developed by Shin and Xiu as a hyper-reduction method. The selection algorithm was originally proposed as a stochastic collocation method for uncertainty quantification. Since the algorithm aims at maximizing a quantity $\mathcal{S}$ that measures both the column orthogonality and the determinant, we refer to the algorithm as S-OPT. Numerical examples are provided to demonstrate the performance of S-OPT and to compare its performance with a gappy proper orthogonal decomposition (POD) algorithm. Here, we found that using the S-OPT algorithm is shown to predict the full order solutions with higher accuracy than gappy POD especially when the number of sampling points is small, although we note that S-OPT shows slow asymptotic convergence with respect to the number of samples for some applications, e.g., Lagrangian hydrodynamics.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS1976

A model for the simulation of atmospheric turbulence

A method that produces realistic simulations of atmospheric turbulence is developed and analyzed. The procedure makes use of a generalized spectral analysis, often called a proper orthogonal decomposition or the Karhunen-Loeve expansion. A set of criteria, emphasizing a realistic appearance, a correct spectral shape, and non-Gaussian statistics, is selected in order to evaluate the model turbulence. An actual turbulence record is analyzed in detail, providing both a standard for comparison and input statistics for the generalized spectral analysis. The simulation utilizes the eigenfunction-expansion procedure to produce preliminary time histories of the three velocity components simultaneously. It is found that important cross-statistical features are reasonably well-behaved.

Petersen, E. L.↗

NASA NTRS1981

On the OLP prediction of the unstable modes of the flat plate turbulent boundary layer

The Orr et al. (OLP) stability analysis is reformulated to yield profiles from hot-wire velocity covariance data in the boundary layer of a flow over a flat plate. The proper orthogonal decomposition theorem (PODT) is used to maximize the mean square of the inner product of the velocity vector with a deterministic candidate vector function in Hilbert space. The OLP formulation couples the PODT for the eddy viscosity with a global extremum principle for the disturbance kinetic energy. Derivation of the governing equations is described analytically, as are the extraction of the mean velocity profile and the eddy viscosity. Finally, techniques for characterizing the dominant mode of the velocity field are introduced and applied to three-dimensional flow over a flat plate. Good agreement is found between the predictions, using laboratory data, and the measured energy and three-dimensional structures of the dominant eddies.

Wadia, A. R.↗

NASA NTRS1985

Large eddy interactions in a turbulent channel flow

The dynamic processes of large eddies in a turbulent channel flow have been examined by utilizing an orthogonal expansion of the velocity fluctuation, known in the literature as the Proper Orthogonal Decomposition Theorem. The mathematical form of these functions is unknown in contrast to the Fourier analysis. Attention is focused on the nonlinear, turbulence-turbulence interaction process in the dynamical equation for large eddies (the first term in the expansion). The nonlinear interactions of the components of the first mode are treated exactly, but influences of higher modes are modeled. This requires adjustment of both the skewness and the effective Reynolds number so that the energy equilibrium of the large eddies is ensured when the mean velocity distribution is assumed known for experiments. Computational results show that the first mode contributes significantly to turbulent intensities and possesses a structural and statistical character similar to that of the entire flow.

Hong, S. K.↗

NASA NTRS1987

A low dimensional dynamical system for the wall layer

Low dimensional dynamical systems which model a fully developed turbulent wall layer were derived.The model is based on the optimally fast convergent proper orthogonal decomposition, or Karhunen-Loeve expansion. This decomposition provides a set of eigenfunctions which are derived from the autocorrelation tensor at zero time lag. Via Galerkin projection, low dimensional sets of ordinary differential equations in time, for the coefficients of the expansion, were derived from the Navier-Stokes equations. The energy loss to the unresolved modes was modeled by an eddy viscosity representation, analogous to Heisenberg's spectral model. A set of eigenfunctions and eigenvalues were obtained from direct numerical simulation of a plane channel at a Reynolds number of 6600, based on the mean centerline velocity and the channel width flow and compared with previous work done by Herzog. Using the new eigenvalues and eigenfunctions, a new ten dimensional set of ordinary differential equations were derived using five non-zero cross-stream Fourier modes with a periodic length of 377 wall units. The dynamical system was integrated for a range of the eddy viscosity prameter alpha. This work is encouraging.

Aubry, N.↗

NASA NTRS1990

Plane waves and structures in turbulent channel flow

A direct simulation of turbulent flow in a channel is analyzed by the method of empirical eigenfunctions (Karhunen-Loeve procedure, proper orthogonal decomposition). This analysis reveals the presence of propagating plane waves in the turbulent flow. The velocity of propagation is determined by the flow velocity at the location of maximal Reynolds stress. The analysis further suggests that the interaction of these waves appears to be essential to the local production of turbulence via bursting or sweeping events in the turbulent boundary layer, with the additional suggestion that the fast acting plane waves act as triggers.

Sirovich, L.↗

NASA NTRS1990

Statistical analysis of near-wall structures in turbulent channel flow

The two-point correlation tensor is used to analyze near-wall structures in turbulent channel flow. Simulation results have been utilized to obtain the correlation tensor with sufficient spatial resolution to permit accurate differentiation for computing the vorticity correlation, and the resolution of sharp features such as the shear layer and its corresponding region of spanwise vorticity. Proper orthogonal decomposition and stochastic estimation were observed to yield similar results when appropriate conditions were employed for the estimation. Resulting structures were such that they would be detected by standard conditional sampling methods. The results of decomposing the vorticity field and the velocity field were significantly different, indicating that if a coherent structure is found that dominates the velocity fields, the curl of that velocity structure will not dominate the vorticity field.

Moser, Robert D.↗

NASA NTRS1989

Low dimensional models of the wall region of a turbulent boundary layer, and the possibility of control

A method of identification of coherent structures in a random turbulent flow is outlined with emphasis on the structure best correlated with a random energy-integrable field and application of the proper orthogonal decomposition to the shear flow of the wall region. The wall region of a pipe flow with almost pure glycerine is investigated experimentally, dynamic equations are obtained, and an energy-transfer model is employed. Numerical solutions are obtained, implications for the flow in the wall region are discussed with focus placed on the sets of solutions exhibiting intermittency. The most significant finding of the study is considered to be the suggestion of the etiology of the bursting phenomenon; presuming that abrupt transitions from one fixed point to the other can be identified with a burst, these bursts appear to be produced autonomously by the wall region, but be triggered by pressure signals from the outer layer.

Lumley, John L.↗

NASA NTRS1992

Application of multipoint measurements for flow characterization

Some of the possibilities for inferring the structure of complicated flows from simultaneous measurements at many points are reviewed. Conditional sampling, pseudoflow visualization, stochastic estimation, and the proper orthogonal decomposition are briefly reviewed and illustrated by example. Resolution criteria for multipoint spatial arrays are proposed that minimize the possibilities for misinterpreting the data.

Glauser, Mark N.↗

NASA NTRS1992

Studies in turbulence

Various papers on turbulence are presented. Individual topics addressed include: modeling the dissipation rate in rotating turbulent flows, mapping closures for turbulent mixing and reaction, understanding turbulence in vortex dynamics, models for the structure and dynamics of near-wall turbulence, complexity of turbulence near a wall, proper orthogonal decomposition, propagating structures in wall-bounded turbulence flows. Also discussed are: constitutive relation in compressible turbulence, compressible turbulence and shock waves, direct simulation of compressible turbulence in a shear flow, structural genesis in wall-bounded turbulence flows, vortex lattice structure of turbulent shear slows, etiology of shear layer vortices, trilinear coordinates in fluid mechanics.

Gatski, Thomas B.↗

NASA NTRS1992

Lumley's PODT definition of large eddies and a trio of numerical procedures

Lumley's 1967 Moscow paper provided, for the first time, a completely rational definition of the physically-useful term 'large eddy', popular for a half-century. The numerical procedures based upon his results are: (1) PODT (Proper Orthogonal Decomposition Theorem), which extracts the Large Eddy structure of stochastic processes from physical or computer simulation two-point covariances, and 2) LEIM (Large-Eddy Interaction Model), a predictive scheme for the dynamical large eddies based upon higher order turbulence modeling. Earlier Lumley's work (1964) forms the basis for the final member of the triad of numerical procedures: this predicts the global neutral modes of turbulence which have surprising agreement with both structural eigenmodes and those obtained from the dynamical equations. The ultimate goal of improved engineering design tools for turbulence may be near at hand, partly due to the power and storage of 'supermicrocomputer' workstations finally becoming adequate for the demanding numerics of these procedures.

Payne, Fred R.↗

NASA NTRS1996

Low-Dimensional Dynamical Models of Thermal Convection

A low-dimensional dynamic model for transitional buoyancy-driven flow in a differentially heated tall enclosure is presented. The full governing partial differential equations with the associated boundary conditions are solved by a spectral element method for a cavity of aspect ratio A=20. Proper orthogonal decomposition is applied to the oscillatory solution at Prandtl number Pr=P tau (omega) = 0.71 and Grashof number G tau (omega) = 3.2 x 10 (exp 4) to construct empirical eigenfunctions. Using the four most energetic empirical eigenfunctions for the velocity and temperature as basis functions and applying Galerkin's method, a reduced model consisting of eight nonlinear ordinary differential equations is obtained. Close to the 'design' conditions (P tau(omega) G tau(omega)), the low-order model (LOM) predictions are in excellent agreement with the predictions of the full model. In particular, the critical Grashof number at the onset of the first temporal flow instability (Hopf bifurcation) was well as the frequency and amplitude of oscillations at supercritical conditions are in excellent agreement with the predictions of the full model. Far from the 'design' conditions, the LOM predicts the existence of multiple stable steady solutions at large values of G tau, and a unique stable steady solution at small values of G tau, and exhibits hysteretic behavior that is qualitatively similar to that observed in direct numerical simulations based on the full model.

Liakopoulos, Anthony↗

NASA NTRS1996

Structure Identification Within a Transitioning Swept-Wing Boundary Layer

Extensive measurements are made in a transitioning swept-wing boundary layer using hot-film, hot-wire and cross-wire anemometry. The crossflow-dominated flow contains stationary vortices that breakdown near mid-chord. The most amplified vortex wavelength is forced by the use of artificial roughness elements near the leading edge. Two-component velocity and spanwise surface shear-stress correlation measurements are made at two constant chord locations, before and after transition. Streamwise surface shear stresses are also measured through the entire transition region. Correlation techniques are used to identify stationary structures in the laminar regime and coherent structures in the turbulent regime. Basic techniques include observation of the spatial correlations and the spatially distributed auto-spectra. The primary and secondary instability mechanisms are identified in the spectra in all measured fields. The primary mechanism is seen to grow, cause transition and produce large-scale turbulence. The secondary mechanism grows through the entire transition region and produces the small-scale turbulence. Advanced techniques use Linear Stochastic Estimation (LSE) and Proper Orthogonal Decomposition (POD) to identify the spatio-temporal evolutions of structures in the boundary layer. LSE is used to estimate the instantaneous velocity fields using temporal data from just two spatial locations and the spatial correlations. Reference locations are selected using maximum RMS values to provide the best available estimates. POD is used to objectively determine modes characteristic of the measured flow based on energy. The stationary vortices are identified in the first laminar modes of each velocity component and shear component. Experimental evidence suggests that neighboring vortices interact and produce large coherent structures with spanwise periodicity at double the stationary vortex wavelength. An objective transition region detection method is developed using streamwise spatial POD solutions which isolate the growth of the primary and secondary instability mechanisms in the first and second modes, respectively. Temporal evolutions of dominant POD modes in all measured fields are calculated. These scalar POD coefficients contain the integrated characteristics of the entire field, greatly reducing the amount of data to characterize the instantaneous field. These modes may then be used to train future flow control algorithms based on neural networks.

Chapman, Keith↗

NASA NTRS1999

Jet Mixing Enhancement by Feedback Control

The objective of this work has been to produce methodologies for high speed jet noise reduction based on natural mechanisms and enhanced feedback control to affect frequencies and structures in a prescribed manner. In this effort the two-point hot wire measurements obtained in the Langley jet facility by Ukeiley were used in conjuction with linear stochastic estimation (LSE) to implement the LSE component of the complementary technique. This method combines the Proper Orthogonal Decomposition (POD) and LSE to provide an experimental low dimensional time dependent description of the flow field. From such a description it should be possible to identify short time high strain rate events in the jet which contribute to the noise. The main task completed for this effort is summarized: LSE experiments were performed at the downstream locations where the two point hot wire measurements have been obtained by Ukeiley. These experiments involved sampling simultaneously hot wire signals from a relatively course spatial grid in gamma and theta. From this simultaneous data, coupled with the two-point measurements of Ukeiley via the LSE components of the complementary technique, an experimental low dimensional description of the jet at 4, 5, 6, 7 and 8 diameters downstream was obtained for Mach numbers of 0.3 and 0.6. We first present an overview of the theory involved. We finish up with a statement of the work performed and finally provide charts from a 1999 APS talk which summarizes the results.

Glauser, Mark↗

NASA NTRS2001

Reduced-Order Modeling: New Approaches for Computational Physics

In this paper, we review the development of new reduced-order modeling techniques and discuss their applicability to various problems in computational physics. Emphasis is given to methods ba'sed on Volterra series representations and the proper orthogonal decomposition. Results are reported for different nonlinear systems to provide clear examples of the construction and use of reduced-order models, particularly in the multi-disciplinary field of computational aeroelasticity. Unsteady aerodynamic and aeroelastic behaviors of two- dimensional and three-dimensional geometries are described. Large increases in computational efficiency are obtained through the use of reduced-order models, thereby justifying the initial computational expense of constructing these models and inotivatim,- their use for multi-disciplinary design analysis.

Beran, Philip S.↗

NASA NTRS2001

System Identification and POD Method Applied to Unsteady Aerodynamics

The representation of unsteady aerodynamic flow fields in terms of global aerodynamic modes has proven to be a useful method for reducing the size of the aerodynamic model over those representations that use local variables at discrete grid points in the flow field. Eigenmodes and Proper Orthogonal Decomposition (POD) modes have been used for this purpose with good effect. This suggests that system identification models may also be used to represent the aerodynamic flow field. Implicit in the use of a systems identification technique is the notion that a relative small state space model can be useful in describing a dynamical system. The POD model is first used to show that indeed a reduced order model can be obtained from a much larger numerical aerodynamical model (the vortex lattice method is used for illustrative purposes) and the results from the POD and the system identification methods are then compared. For the example considered, the two methods are shown to give comparable results in terms of accuracy and reduced model size. The advantages and limitations of each approach are briefly discussed. Both appear promising and complementary in their characteristics.

Tang, Deman↗

NASA NTRS2003

Low Dimensional Tools for Flow-Structure Interaction Problems: Application to Micro Air Vehicles

A low dimensional tool for flow-structure interaction problems based on Proper Orthogonal Decomposition (POD) and modified Linear Stochastic Estimation (mLSE) has been proposed and was applied to a Micro Air Vehicle (MAV) wing. The method utilizes the dynamic strain measurements from the wing to estimate the POD expansion coefficients from which an estimation of the velocity in the wake can be obtained. For this experiment the MAV wing was set at five different angles of attack, from 0 deg to 20 deg. The tunnel velocities varied from 44 to 58 ft/sec with corresponding Reynolds numbers of 46,000 to 70,000. A stereo Particle Image Velocimetry (PIV) system was used to measure the wake of the MAV wing simultaneously with the signals from the twelve dynamic strain gauges mounted on the wing. With 20 out of 2400 POD modes, a reasonable estimation of the flow flow was observed. By increasing the number of POD modes, a better estimation of the flow field will occur. Utilizing the simultaneously sampled strain gauges and flow field measurements in conjunction with mLSE, an estimation of the flow field with lower energy modes is reasonable. With these results, the methodology for estimating the wake flow field from just dynamic strain gauges is validated.

Schmit, Ryan F.↗