Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Reduced Order Modeling”

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.

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

At least 595 records · Page 33
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Nonlinear proper orthogonal decomposition for convection-dominated flows

Autoencoder techniques find increasingly common use in reduced order modeling as a means to create a latent space. This reduced order representation offers a modular data-driven modeling approach for nonlinear dynamical systems when integrated with a time series predictive model. In this Letter, we put forth a nonlinear proper orthogonal decomposition (POD) framework, which is an end-to-end Galerkin-free model combining autoencoders with long short-term memory networks for dynamics. By eliminating the projection error due to the truncation of Galerkin models, a key enabler of the proposed nonintrusive approach is the kinematic construction of a nonlinear mapping between the full-rank expansion of the POD coefficients and the latent space where the dynamics evolve. We test our framework for model reduction of a convection-dominated system, which is generally challenging for reduced order models. Our approach not only improves the accuracy, but also significantly reduces the computational cost of training and testing.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2016

Applicability of aquifer impact models to support decisions at CO 2 sequestration sites

The National Risk Assessment Partnership has developed a suite of tools to assess and manage risk at CO 2 sequestration sites. This capability includes polynomial or look-up table based reduced-order models (ROMs) that predict the impact of CO 2 and brine leaks on overlying aquifers. The development of these computationally-efficient models and the underlying reactive transport simulations they emulate has been documented elsewhere (Carroll et al., 2014a; Carroll et al., 2014b; Dai et al., 2014 ; Keating et al., 2016). Here in this paper, we seek to demonstrate applicability of ROM-based analysis by considering what types of decisions and aquifer types would benefit from the ROM analysis. We present four hypothetical examples where applying ROMs, in ensemble mode, could support decisions during a geologic CO 2 sequestration project. These decisions pertain to site selection, site characterization, monitoring network evaluation, and health impacts. In all cases, we consider potential brine/CO 2 leak rates at the base of the aquifer to be uncertain. We show that derived probabilities provide information relevant to the decision at hand. Although the ROMs were developed using site-specific data from two aquifers (High Plains and Edwards), the models accept aquifer characteristics as variable inputs and so they may have more broad applicability. We conclude that pH and TDS predictions are the most transferable to other aquifers based on the analysis of the nine water quality metrics (pH, TDS, 4 trace metals, 3 organic compounds). Guidelines are presented for determining the aquifer types for which the ROMs should be applicable.

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS1992

On-line implementation of nonlinear parameter estimation for the Space Shuttle main engine

We investigate the performance of a nonlinear estimation scheme applied to the estimation of several parameters in a performance model of the Space Shuttle Main Engine. The nonlinear estimator is based upon the extended Kalman filter which has been augmented to provide estimates of several key performance variables. The estimated parameters are directly related to the efficiency of both the low pressure and high pressure fuel turbopumps. Decreases in the parameter estimates may be interpreted as degradations in turbine and/or pump efficiencies which can be useful measures for an online health monitoring algorithm. This paper extends previous work which has focused on off-line parameter estimation by investigating the filter's on-line potential from a computational standpoint. ln addition, we examine the robustness of the algorithm to unmodeled dynamics. The filter uses a reduced-order model of the engine that includes only fuel-side dynamics. The on-line results produced during this study are comparable to off-line results generated previously. The results show that the parameter estimates are sensitive to dynamics not included in the filter model. Off-line results using an extended Kalman filter with a full order engine model to address the robustness problems of the reduced-order model are also presented.

Buckland, Julia H.↗
NASA NTRSDate not supplied

An Interactive MATLAB Program for Fitting Transfer Functions to Frequency Responses

A computer program called FRFit (Frequency Response Fitting) is described for fitting single-input single-output transfer function models to empirical frequency response data. The program is interactive in that the user specifies ``elementary factors'' (gain, delay, pure differentiators and integrators, and first- and second-order zeros and poles) by entering numerical values or moving sliders in a graphical user interface. A nonlinear optimization can then be performed to obtain maximum likelihood estimates of transfer function parameters and uncertainties to provide feedback on the modeling and refine estimates. Several examples are discussed, including the identification of aircraft pitch dynamics from simulation data and data reported in the literature, approximating Theodorsen's function of unsteady aerodynamics, and obtaining a reduced-order model of a computational fluid dynamics code describing the unsteady aerodynamics around an aeroelastic wing. The program has some usefulness as a teaching aid, and can be applied to model structure determination, reduced-order modeling, preliminary analysis, and simple system identification problems. The program was written in MATLAB and is planned for public release through the NASA Software Catalog.

System identification↗
DOE OSTI2022

Variational encoder geostatistical analysis (VEGAS) with an application to large scale riverine bathymetry

Estimation of riverbed profiles, also known as bathymetry, plays a vital role in many applications, such as safe and efficient inland navigation, prediction of bank erosion, land subsidence, and flood risk management. The high cost and complex logistics of direct bathymetry surveys, i.e, depth imaging, have encouraged the use of indirect measurements such as surface flow velocities. However, estimating high-resolution bathymetry from indirect measurements is an inverse problem that can be computationally challenging. Here, we propose a reduced-order model (ROM) based approach that utilizes a variational autoencoder (VAE), a type of deep neural network with a narrow layer in the middle, to compress bathymetry and flow velocity information and accelerate bathymetry inverse problems from flow velocity measurements. In our application, the shallow-water equations (SWE) with appropriate boundary conditions (BCs), e.g., the discharge and/or the free surface elevation, constitute the forward problem, to predict flow velocity. Then, ROMs of the SWEs are constructed on a nonlinear manifold of low dimensionality through a variational encoder and the bathymetry inversion problem is derived on the low-dimensional latent space in a Hierarchical Bayesian setting. Further, the reformulation allows variational inference with a small number (e.g., $\mathscr{O}$ (100) of ROM runs and efficient uncertainty quantification. We have tested our inversion approach on a one-mile reach of the Savannah River, GA, USA. Once the neural network is trained (offline stage), the proposed technique can perform the inversion operation orders of magnitude faster than traditional inversion methods that are commonly based on linear projections, such as principal component analysis (PCA), or the principal component geostatistical approach (PCGA). Furthermore, tests show that the algorithm can estimate the bathymetry with good accuracy even with sparse flow velocity measurements.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2024

pvcracks

SAND2024-00922O This software uses electroluminescence images to predict power loss due to cell cracks in photovoltaic modules. The software will incorporate trained variational autoencoder(s) to parameterize cell cracks detected in electroluminescence images of photovoltaic modules and relate cracks to electrical properties; reduced-order models of finite element simulations of electrical behavior of photovoltaic modules under thermomechanical stresses; reduced-order models of stress distributions; inside photovoltaic modules resulting from x-ray topography experiments; and image segmentation and splining methods to analyze x-ray topography measurements of cracked photovoltaic cells. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.

Hartley, James↗
DOE OSTI2023

gLaSDI: Parametric physics-informed greedy latent space dynamics identification

A parametric adaptive physics-informed greedy Latent Space Dynamics Identification (gLaSDI) method is proposed for accurate, efficient, and robust data-driven reduced-order modeling of high-dimensional nonlinear dynamical systems. In the proposed gLaSDI framework, an autoencoder discovers intrinsic nonlinear latent representations of high-dimensional data, while dynamics identification (DI) models capture local latent-space dynamics. Here, an interactive training algorithm is adopted for the autoencoder and local DI models, which enables identification of simple latent-space dynamics and enhances accuracy and efficiency of data-driven reduced-order modeling. To maximize and accelerate the exploration of the parameter space for the optimal model performance, an adaptive greedy sampling algorithm integrated with a physics-informed residual-based error indicator and random-subset evaluation is introduced to search for the optimal training samples on the fly. Further, to exploit local latent-space dynamics captured by the local DI models for an improved modeling accuracy with a minimum number of local DI models in the parameter space, a -nearest neighbor convex interpolation scheme is employed. The effectiveness of the proposed framework is demonstrated by modeling various nonlinear dynamical problems, including Burgers equations, nonlinear heat conduction, and radial advection. The proposed adaptive greedy sampling outperforms the conventional predefined uniform sampling in terms of accuracy. Compared with the high-fidelity models, gLaSDI achieves 17 to 2,658× speed-up with 1 to 5% relative errors.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1990

Reduced order component models for flexible multibody dynamics simulations

Many flexible multibody dynamics simulation codes require some form of component description that properly characterizes the dynamic behavior of the system. A model reduction procedure for producing low order component models for flexible multibody simulation is described. Referred to as projection and assembly, the method is a Rayleigh-Ritz approach that uses partitions of the system modal matrix as component Ritz transformation matrices. It is shown that the projection and assembly method yields a reduced system model that preserves a specified set of the full order system modes. Unlike classical component mode synthesis methods, the exactness of the method described is obtained at the expense of having to compute the full order system modes. The paper provides a comprehensive description of the method, a proof of exactness, and numerical results demonstrating the method's effectiveness.

Tsuha, Walter S.↗
NASA NTRS1978

Procedures for generation and reduction of linear models of a turbofan engine

A real time hybrid simulation of the Pratt & Whitney F100-PW-F100 turbofan engine was used for linear-model generation. The linear models were used to analyze the effect of disturbances about an operating point on the dynamic performance of the engine. A procedure that disturbs, samples, and records the state and control variables was developed. For large systems, such as the F100 engine, the state vector is large and may contain high-frequency information not required for control. This, reducing the full-state to a reduced-order model may be a practicable approach to simplifying the control design. A reduction technique was developed to generate reduced-order models. Selected linear and nonlinear output responses to exhaust-nozzle area and main-burner fuel flow disturbances are presented for comparison.

Seldner, K.↗
DOE OSTI2016

The Reducibility of Reactor Physics Depletion and Decay Calculation

Depletion and decay calculations are integral part of reactor physics calculations, necessary to estimate fuel characteristics following a given exposure to radiation. Given the complexity of reactor design and the unprecedented sophistication of the new high fidelity modeling and simulation tools, attempting to calculate a very accurate estimation of fuel characteristics on a pin-by-pin level, there is an ever-increasing need to perform these calculations in a computationally efficient manner. This summary presents a new enabling methodology to achieve higher computational gains while addressing one of the fundamental challenges of existing methods, that is how to preserve the accuracy of the original high fidelity models. This is difficult with current approaches relying on look-up tables or the elimination of non-influential isotopes. To achieve that we employ a reduced order modeling approach which reduces the effective dimensionality of the solution space, represented by the number of tracked nuclides. This approach allows the analyst to devise a user-defined level of reduction supported by rigorous error bounds that quantify the accuracy of the reduced model. This summary explores this new methodology by calculating the effective dimensionality for representative BWR depletion problems that are required to meet user-defined accuracy for the nuclides concentrations. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2024

Performance characterization and modeling of an oscillating surge wave energy converter

Abstract Testing wave energy converters in the ocean could be expensive and complex, which necessitates the use of numerical modeling. However, accurately modeling the response of wave energy converters with high-fidelity simulations can be computationally intensive in the design stage where different configurations must be considered. Reduced-order models based on simplified equations of motion can be very useful in the design, optimization, or control of wave energy converters. Given the complex dynamics of wave energy converters, accurate representation, and evaluation of relative contributions by different forces are required. This effort is concerned with a performance characterization of the hydrodynamic response of an oscillating surge wave energy converter that is based on a reduced-order model. A state-space model is used to represent the radiation damping term. Morison’s representation of unsteady forces is used to account for the nonlinear damping. Wave tank tests are performed to validate simulations. A free response simulation is used to determine the coefficients of the state-space model. Torque-forced simulations are used to identify the coefficients of the nonlinear damping term for different amplitudes and wave frequencies. The impact of varying these coefficients on the response is investigated. An assessment of the capability of the model in predicting the hydrodynamic response under irregular forcing is performed. The results show that the maximum error is 3% when compared with high-fidelity simulations. It is determined that the nonlinear damping is proportional to the torque amplitude and its effects are more pronounced as the amplitude of the flap oscillations increases.

16 TIDAL AND WAVE POWER↗
DOE OSTI2024

Micro–macro finite element modeling method for rub response in abradable coating materials

Gas turbine engines experience “rub” when the rotating blades come in contact with a static abradable coating. This results in extreme strain rates and dynamics inside a high-temperature/high-pressure environment. Current rub models are phenomenological and do not reflect the underlying microstructures, thus limiting their prediction accuracy. In this work, a microstructure-informed, reduced order modeling framework is introduced for simulating abradable coating “rub" behavior. This framework comprises a microscale model constructed based on digitized abradable microstructure and explicitly simulates the mechanical behavior of each constituent phases and their interactions. After calibration and validation with experiment data, the calibrated microscale model is used to generate data across a vast range of applied strain rates and temperature with various load paths. Then, the virtually generated data are used to fit the macroscopic-reduced order model, which enables fast component scale rub simulation without compromising the integrity of the complex material behavior. In conclusion, the proposed effort will address the technical challenge of predicting abradable material behavior during rub through the application of multiscale modeling from microstructure to engines behavior, effectively reducing the development costs and time of new abradable material for better “rub” properties.

36 MATERIALS SCIENCE↗
DOE OSTI2021

Physics-Informed Machine Learning Models for Predicting the Progress of Reactive-Mixing

This paper presents a physics-informed machine learning (ML) framework to construct reduced-order models (ROMs) for reactive-transport quantities of interest (QoIs) based on high-fidelity numerical simu-lations. QoIs include species decay, product yield, and degree of mixing. The ROMs for QoIs are applied to quantify and understand how the chemical species evolve over time. First, high-resolution datasets for constructing ROMs are generated by solving anisotropic reaction-di?usion equations using a non-negative finite element formulation for di?erent input parameters. The reactive-mixing model input parameters are: time-scale associated with flipping of velocity, spatial-scale controlling small/large vortex structures of velocity, perturbation parameter of the vortex-based velocity, anisotropic dispersion strength/contrast, and molecular diffusion. Second, random forests, F-test, and mutual information criterion are used to evaluate the importance of model inputs/features with respect to QoIs. We observed that anisotropic dispersion strength/contrast is the most important feature and time-scale associated with flipping of velocity is the least important feature. Third, Support Vector Machines (SVM) and Support Vector Regression (SVR) are used to construct ROMs based on the model inputs. The constructed SVR-ROMs are then used to predict scaling of QoIs. We also present estimates and inequalities on the QoIs, which inform that the species decay, mix, and produce in an exponential fashion. These inequalities also inform that a radial basis function is the most suitable kernel for the SVM/SVR models for QoIs. It is observed that R2-score for SVR-ROMs on unseen data is greater than 0.9, implying that the SVR-ROMs are able to predict the reaction-diffusion system state reasonably well. Finally, in terms of the computational cost, the proposed SVM-ROMs are O(107) times faster than running a high-fidelity finite element simulation for evaluating QoIs. This makes the proposed ML-based ROMs attractive for reactive-transport sensing and real-time monitoring applications as they are significantly faster yet reasonably accurate.

Mudunuru, Maruti K.↗
DOE OSTI2025

Develop a new integrated macro→micro←nano (MMN) multiscale modeling framework to optimize high strength aluminum alloys and processes for vehicle light-weighting​

Bending tests provide a means to study plane strain fracture performance of 6000 series aluminum alloys. Metrics from bending tests have been correlated with self-pierce riveting (SPR) performance of a high strength AA6111 automotive aluminum alloy in previous works. Using the ORNL HPC resources, this project developed an innovative macro→micro←nano (MMN) multiscale microstructure-based finite element (FE) code to further understanding of the relationship between microstructure and fracture properties of high-strength 6000-series alloys. This work started with microstructural characterization in both mesoscale and nanoscale and bending performance characterization of AA6111 HS2-T6 alloy at Ford, the MMN framework was applied to this alloy to simulate 3-point VDA bending. From the results of the macro-modeling of 3-point VDA bending, the critical region of fracture was identified, and the region geometry was used to construct the micro-model. The fracture criterion of micron-scale precipitates and aluminum matrix which contains submicron and nano particles (AL-SMP-NP) in the micro-model was calibrated and validated by comparing simulated and measured bending results. With the AL-SMP-NP fracture strain obtained, the fracture strain of Al-matrix containing nano particles (ALNP) will similarly be determined by a submicron scale-model using an edge-constrained FE modeling approach developed by Hu et al. With the ALNP fracture strain obtained, the fracture strain of Al-matrix containing no particles will similarly be determined by a nano scale-model using an edge-constrained FE modeling approach. After the MMN framework is built and fracture criterion calibrated, nano-model FE simulations with virtual microstructures was performed to obtain a reduced order model (ROM) of the fracture criterion of the ALNP as a function of volume fraction, size, and distribution of the nanoparticle. This nano→submicron→micro modeling part allows exploration of the influence of different material nanostructures from different process conditions on the bending properties within the multiscale bending simulation framework and the ROM of material bendability as a function of nanoparticle size and shape was established. This obtained reduced order model (ROM) could help guide the design and selection of materials to improve existing SPR process models that could replace trial-and-error rivet/die selection and help to design new rivet/die combinations capable of robustly joining new higher strength 6000 5 series alloys in automotive body structures. This would enable lightweighting of Ford vehicles leading to greater fuel efficiency and reduce manufacturing time and energy.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Validation and parameterization of a novel physics-constrained neural dynamics model applied to turbulent fluid flow

We report, in fluid physics, data-driven models to enhance or accelerate time to solution are becoming increasingly popular for many application domains, such as alternatives to turbulence closures, system surrogates, or for new physics discovery. In the context of reduced order models of high-dimensional time-dependent fluid systems, machine learning methods grant the benefit of automated learning from data, but the burden of a model lies on its reduced-order representation of both the fluid state and physical dynamics. In this work, we build a physics-constrained, data-driven reduced order model for Navier–Stokes equations to approximate spatiotemporal fluid dynamics in the canonical case of isotropic turbulence in a triply periodic box. The model design choices mimic numerical and physical constraints by, for example, implicitly enforcing the incompressibility constraint and utilizing continuous neural ordinary differential equations for tracking the evolution of the governing differential equation. We demonstrate this technique on a three-dimensional, moderate Reynolds number turbulent fluid flow. In assessing the statistical quality and characteristics of the machine-learned model through rigorous diagnostic tests, we find that our model is capable of reconstructing the dynamics of the flow over large integral timescales, favoring accuracy at the larger length scales. More significantly, comprehensive diagnostics suggest that physically interpretable model parameters, corresponding to the representations of the fluid state and dynamics, have attributable and quantifiable impact on the quality of the model predictions and computational complexity.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2026

Scientific Discovery with Physics-Informed System Identification (Abbreviated Report)

My fellowship research focused on making physics-based simulations faster and more useful through machine learning. Many problems in science and engineering are governed by partial differential equations, but high-fidelity simulations are often too expensive to run repeatedly. I worked on improving Latent Space Dynamics Identification (LaSDI), a reduced-order modeling framework that compresses large simulation data sets into a smaller representation and then learns how that representation evolves over time. The motivation was to develop reduced models that remain accurate for more challenging systems, especially when predictions must remain reliable over long time intervals or when the underlying dynamics are more complicated than standard methods can easily handle. I also contributed to related work on Quandary, a high-performance software effort for simulation and control of open quantum systems, before focusing primarily on Latent Space Dynamics Identification methods. The main outcomes of the fellowship were two new algorithms (both of which were published), Rollout-LaSDI and Higher-Order LaSDI, together with supporting work on multi-stage Latent Space Dynamics Identification. Rollout-LaSDI improved long-term prediction by training the model to stay accurate over extended time horizons, and Higher-Order LaSDI broadened the method so it could model systems with higher-order time dynamics. My contributions to multistage Latent Space Dynamics Identification also helped show that its later training stages could be simplified without losing effectiveness, and that this behavior held across different model architectures and training strategies. Taken together, these advances improved the accuracy, flexibility, and practical value of reduced-order modeling tools for computational science.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Enabling Dynamic Probabilistic Risk Assessment of Physical Security Using EMRALD and MAAP

The optimization of physical security in nuclear power plants requires sophisticated methodologies that integrate operator actions and plant behavior through advanced simulation tools. To address this, Idaho National Laboratory [JL2.1]has developed the Modeling and Analysis for Safety and Security using the Dynamic EMRALD Framework (MASS-DEF) methodology, an approach that integrates force-on-force simulations, dynamic probabilistic risk assessment, and thermal-hydraulics modeling [JL3.1]to enhance security planning while reducing costs. We developed a tool that produces reduced order models using thermal hydraulic simulations from the Modular Accident Analysis Program (MAAP) [1]. These models can quickly evaluate reactor core behavior during attack simulations, and in so doing, address two barriers of traditional methods: (1) MAAP simulations are computationally intensive, and (2) attack scenarios must be run in a secure environment, which complicates analysis and validation. By precomputing scenario outcomes for a small number of modified parameters, the reduced order model significantly decreases the computational cost and enables offsite review of the results.

22 - GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2022

Parametric model-order reduction for radiation transport using multi-resolution proper orthogonal decomposition

For parametric high-fidelity simulations, it is often desirable to utilize a reduced-order model (ROM) to emulate, at a reduced computational cost, parametric solutions of the governing partial differential equations (PDEs) for unseen parameter values. One commonly employed option is to utilize a data-driven, projection-based ROM supplemented with subspace identification via proper orthogonal decomposition (POD). POD discovers the ROM subspace by computing the singular value decomposition (SVD) of a set of training data from the full-order model (FOM). In streaming-dominated radiation transport simulations with localized sources, solutions often greatly vary over the spatial domain by many orders of magnitude. In such cases, machine-precision arithmetic can be insufficient to obtain an accurate SVD, resulting in a poorly performing ROM. We present a method called multiresolution POD (mrPOD) that mitigates these inaccuracies. The mrPOD method works by decomposing the spatial domain into regions and performing proper orthogonal decomposition on the training dataset separately in each region. In conclusion, mrPOD is tested on single energy group and multigroup atmospheric shielding transport problems and is shown to outperform classic POD.

42 ENGINEERING↗