Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Numerical Representation”

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 1,765 records · Page 98

Learning high-dimensional parametric maps via reduced basis adaptive residual networks

We propose a scalable framework for the learning of high-dimensional parametric maps via adaptively constructed residual network (ResNet) maps between reduced bases of the inputs and outputs. When just few training data are available, it is beneficial to have a compact parametrization in order to ameliorate the ill-posedness of the neural network training problem. By linearly restricting high-dimensional maps to informed reduced bases of the inputs, one can compress high-dimensional maps in a constructive way that can be used to detect appropriate basis ranks, equipped with rigorous error estimates. A scalable neural network learning framework is thus to learn the nonlinear compressed reduced basis mapping. Unlike the reduced basis construction, however, neural network constructions are not guaranteed to reduce errors by adding representation power, making it difficult to achieve good practical performance. Inspired by recent approximation theory that connects ResNets to sequential minimizing flows, we present an adaptive ResNet construction algorithm. This algorithm allows for depth-wise enrichment of the neural network approximation, in a manner that can achieve good practical performance by first training a shallow network and then adapting. We prove universal approximation of the associated neural network class for $L^2_v$ functions on compact sets. Our overall framework allows for constructive means to detect appropriate breadth and depth, and related compact parametrizations of neural networks, significantly reducing the need for architectural hyperparameter tuning. Numerical experiments for parametric PDE problems and a 3D CFD wing design optimization parametric map demonstrate that the proposed methodology can achieve remarkably high accuracy for limited training data, and outperformed other neural network strategies we compared against.

42 ENGINEERING↗

Accelerating Multivariate Functional Approximation Computation with Domain Decomposition Techniques⋆

Modeling large datasets through Multivariate Functional Approximations (MFA) provide an elegant way to handle many visualization and scientific analysis workflows. The process necessitates scalable data partitioning methods to compute MFA representations efficiently without compromising the accuracy or continuity of the reconstructed solution. We propose a domain -decomposed method for computing the MFA with B -spline bases, which reduces the total work per task and uses a restricted Additive Schwarz (RAS) method to converge the control point data degrees -of -freedom along subdomain boundaries. We provide an in-depth analysis of the parallel approach with domain decomposition solvers, aiming to minimize local subdomain error residuals and recover high -order continuity at subdomain interfaces with appropriate choices of knot overlaps. The communication cost, determined by the overlap regions in the RAS implementation, is optimized to recover the numerical error profile of the single subdomain case. Our proposed method stands in contrast to previous methods, which typically only recover either C 0 or at best C 1 continuity for arbitrary B -spline degree expansions, or those that require post -processing to blend discontinuities in the reconstructed data. We demonstrate the effectiveness of our approach using analytical and real -world datasets in 1D, 2D, and 3D through both strong and weak scaling studies. The performance results indicate that the overall cost of computing the approximation is directly proportional to the underlying nearest -neighbor communication implementation, and is only weakly dependent on the overlap region size that determines the size of the messages. This finding underscores the efficiency and scalability of our proposed method, making it a promising solution for handling large datasets in scientific workflows.

additive Schwarz solvers↗

Evaluation of Turbulence and Dispersion in Multiscale Atmospheric Simulations over Complex Urban Terrain during the Joint Urban 2003 Field Campaign

Abstract This paper evaluates the representation of turbulence and its effect on transport and dispersion within multiscale and microscale-only simulations in an urban environment. These simulations, run using the Weather Research and Forecasting Model with the addition of an immersed boundary method, predict transport and mixing during a controlled tracer release from the Joint Urban 2003 field campaign in Oklahoma City, Oklahoma. This work extends the results of a recent study through analysis of turbulence kinetic energy and turbulence spectra and their role in accurately simulating wind speed, direction, and tracer concentration. The significance and role of surface heat fluxes and use of the cell perturbation method in the numerical simulation setup are also examined. Our previous study detailed the model development necessary for our multiscale simulations, examined model skill at predicting wind speeds and tracer concentrations, and demonstrated that dynamic downscaling from mesoscale to microscale through a sequence of nested simulations can improve predictions of transport and dispersion relative to a microscale-only simulation forced by idealized meteorology. Here, predictions are compared with observations to assess qualitative agreement and statistical model skill at predicting wind speed, wind direction, tracer concentration, and turbulent kinetic energy at locations throughout the city. We also investigate the scale distribution of turbulence and the associated impact on model skill, particularly for predictions of transport and dispersion. Our results show that downscaled large-scale turbulence, which is unique to the multiscale simulations, significantly improves predictions of tracer concentrations in this complex urban environment. Significance Statement Simulations of atmospheric transport and mixing in urban environments have many applications, including pollution modeling for urban planning or informing emergency response following a hazardous release. These applications include phenomena with spatial scales spanning from millimeters to kilometers. Most simulations resolve flow only within the urban area of interest, omitting larger scales of turbulence and regional influences. This study examines a method that resolves both the small and large-scale flow features. We evaluate simulation accuracy by comparing predictions with observations from an experiment involving the release of a tracer gas in Oklahoma City, Oklahoma, with emphasis on correctly modeling turbulent fluctuations. Our results demonstrate the importance of resolving large-scale flow features when predicting transport and dispersion in urban environments.

42 ENGINEERING↗

Statistical Distributions for Mesh Independent Solutions in ALEGRA

The representation of material heterogeneity (also referred to as "spatial variation") plays a key role in the material failure simulation method used in ALEGRA. ALEGRA is an arbitrary Lagrangian-Eulerian shock and multiphysics code developed at Sandia National Laboratories and contains several methods for incorporating spatial variation into simulations. A desirable property of a spatial variation method is that it should produce consistent stochastic behavior regardless of the mesh used (a property referred to as "mesh independence"). However, mesh dependence has been reported using the Weibull distribution with ALEGRA's spatial variation method. This report describes efforts towards providing additional insight into both the theory and numerical experiments investigating such mesh dependence. In particular, we have implemented a discrete minimum order statistic model with properties that are theoretically mesh independent.

36 MATERIALS SCIENCE↗

Simulation of Wingtip Vortex Flows with Reynolds-Averaged Navier–Stokes and Scale-Resolving Simulation Methods

The simulation of wingtip vortex flows is reported on with Reynolds-averaged Navier–Stokes (RANS) equations and scale-resolving simulation (SRS) models. These range from turbulent viscosity and Reynolds-stress model (RSM) RANS closures, to hybrid and bridging SRS methods. The aim of the study was threefold: assess the relevance of replicating the experimental flow conditions, evaluate the numerical requisites of each mathematical model, and determine their modeling accuracy in prediction of a representative wingtip vortex flow. The selected problem is the flow around a NACA 0012 wing at 10 deg of angle of attack and Reynolds number of 4.60×10 6 . The results confirm the relevance of reproducing the experimental flow conditions, in particular at the inlet boundary where the flow is not uniform. For this reason, the evaluation of modeling errors using the available measurements requires the specification of the experimental inlet conditions. On the other hand, the quantification of the modeling error indicates that solely RANS–RSM can achieve an accurate representation of the flow dynamics. Whereas the well-recognized limitations of turbulent viscosity RANS closures to deal with solid-body rotation lead to the overprediction of turbulence and consequent rapid diffusion of the wingtip vortex, the predictive use of SRS methods may reveal excessively complex and demanding.

42 ENGINEERING↗

A sharp interface Lagrangian-Eulerian method for flexible-body fluid-structure interaction

This paper introduces a sharp-interface approach to simulating fluid-structure interaction (FSI) involving flexible bodies described by general nonlinear material models and across a broad range of mass density ratios. This new flexible-body immersed Lagrangian-Eulerian (ILE) scheme extends our prior work on integrating partitioned and immersed approaches to rigid-body FSI. Our numerical approach incorporates the geometrical and domain solution flexibility of the immersed boundary (IB) method with an accuracy comparable to body-fitted approaches that sharply resolve flows and stresses up to the fluid-structure interface. Unlike many IB methods, our ILE formulation uses distinct momentum equations for the fluid and solid subregions with a Dirichlet-Neumann coupling strategy that connects fluid and solid subproblems through simple interface conditions. As in earlier work, we use approximate Lagrange multiplier forces to treat the kinematic interface conditions along the fluid-structure interface. This penalty approach simplifies the linear solvers needed by our formulation by introducing two representations of the fluid-structure interface, one that moves with the fluid and another that moves with the structure, that are connected by stiff springs. This approach also enables the use of multi-rate time stepping, which allows us to use different time step sizes for the fluid and structure subproblems. Our fluid solver relies on an immersed interface method (IIM) for discrete surfaces to impose stress jump conditions along complex interfaces while enabling the use of fast structured-grid solvers for the incompressible Navier-Stokes equations. The dynamics of the volumetric structural mesh are determined using a standard finite element approach to large-deformation nonlinear elasticity via a nearly incompressible solid mechanics formulation. This formulation also readily accommodates compressible structures with a constant total volume, and it can handle fully compressible solid structures for cases in which at least part of the solid boundary does not contact the incompressible fluid. Selected grid convergence studies demonstrate second-order convergence in volume conservation and in the pointwise discrepancies between corresponding positions of the two interface representations as well as between first and second-order convergence in the structural displacements. The time stepping scheme is also demonstrated to yield second-order convergence. To assess and validate the robustness and accuracy of the new algorithm, comparisons are made with computational and experimental FSI benchmarks. Test cases include both smooth and sharp geometries in various flow conditions. Furthermore, we also demonstrate the capabilities of this methodology by applying it to model the transport and capture of a geometrically realistic, deformable blood clot in an inferior vena cava filter.

97 MATHEMATICS AND COMPUTING↗

Quantum process identification: a method for characterizing non-markovian quantum dynamics

Established methods for characterizing quantum information processes do not capture non-Markovian (history-dependent) behaviors that occur in real systems. These methods model a quantum process as a fixed map on the state space of a predefined system of interest. Such a map averages over the system's environment, which may retain some effect of its past interactions with the system and thus have a history-dependent influence on the system. Although the theory of non-Markovian quantum dynamics is currently an active area of research, a systematic characterization method based on a general representation of non-Markovian dynamics has been lacking. In this article we present a systematic method for experimentally characterizing the dynamics of open quantum systems. Our method, which we call quantum process identification (QPI), is based on a general theoretical framework which relates the (non-Markovian) evolution of a system over an extended period of time to a time-local (Markovian) process involving the system and an effective environment. In practical terms, QPI uses time-resolved tomographic measurements of a quantum system to construct a dynamical model with as many dynamical variables as are necessary to reproduce the evolution of the system. Through numerical simulations, we demonstrate that QPI can be used to characterize qubit operations with non-Markovian errors arising from realistic dynamics including control drift, coherent leakage, and coherent interaction with material impurities.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Comparison of Joint Modeling Approaches Including Eulerian Sliding Interfaces

Accurate representation of discontinuities such as joints and faults is a key ingredient for high fidelity modeling of shock propagation in geologic media. The following study was done to improve treatment of discontinuities (joints) in the Eulerian hydrocode GEODYN (Lomov and Liu 2005). Lagrangian methods with conforming meshes and explicit inclusion of joints in the geologic model are well suited for such an analysis. Unfortunately, current meshing tools are unable to automatically generate adequate hexahedral meshes for large numbers of irregular polyhedra. Another concern is that joint stiffness in such explicit computations requires significantly reduced time steps, with negative implications for both the efficiency and quality of the numerical solution. An alternative approach is to use non-conforming meshes and embed joint information into regular computational elements. However, once slip displacement on the joints become comparable to the zone size, Lagrangian (even non-conforming) meshes could suffer from tangling and decreased time step problems. The use of non-conforming meshes in an Eulerian solver may alleviate these difficulties and provide a viable numerical approach for modeling the effects of faults on the dynamic response of geologic materials. We studied shock propagation in jointed/faulted media using a Lagrangian and two Eulerian approaches. To investigate the accuracy of this joint treatment the GEODYN calculations have been compared with results from the Lagrangian code GEODYN-L which uses an explicit treatment of joints via common plane contact. We explore two approaches to joint treatment in the code, one for joints with finite thickness and the other for tight joints. In all cases the sliding interfaces are tracked explicitly without homogenization or blending the joint and block response into an average response. In general, rock joints will introduce an increase in normal compliance in addition to a reduction in shear strength. In the present work we consider the limiting case of stiff discontinuities that only affect the shear strength of the material.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Diagnosing the Nature of Land-Atmosphere Coupling: A Case Study of Dry/Wet Extremes

Land-atmosphere (L-A) interactions play a critical role in determining the diurnal evolution of land surface and planetary boundary layer (PBL) temperature and moisture states and fluxes. In turn, these interactions regulate the strength of the connection between surface moisture and precipitation in a coupled system. To address deficiencies in numerical weather prediction and climate models due to improper treatment of L-A interactions, recent studies have focused on development of diagnostics to quantify the strength and accuracy of the land-PBL coupling at the process-level. In this study, a diagnosis of the nature and impacts oflocalland-atmosphere coupling (LoCo) during dry and wet extreme conditions is presented using a combination of models and observations during the summers of2006-7 in the U.S. Southern Great Plains. Specifically, the Weather Research and Forecasting (WRF) model has been coupled to NASA's Land Information System (LIS), which provides a flexible and high-resolution representation and initialization of land surface physics and states. A range of diagnostics exploring the links and feedbacks between soil moisture and precipitation are examined for the dry/wet regimes of this region, along with the behavior and accuracy of different land-PBL scheme couplings under these conditions. In addition, we examine the impact of improved specification ofland surface states, anomalies, and fluxes that are obtained through the use of a hew optimization and uncertainty module in LIS, on the L-A coupling in WRF forecasts. Results demonstrate how LoCo diagnostics can be applied to coupled model components in the context of their integrated impacts on the process-chain connecting the land surface to the PBL and support of hydrological anomalies.

Santanello, Joseph A., Jr.↗

An integral equation formulation for rigid bodies in Stokes flow in three dimensions

Here we present a new derivation of a boundary integral equation (BIE) for simulating the three-dimensional dynamics of arbitrarily-shaped rigid particles of genus zero immersed in a Stokes fluid, on which are prescribed forces and torques. Our method is based on a single-layer representation and leads to a simple second-kind integral equation. It avoids the use of auxiliary sources within each particle that play a role in some classical formulations. We use a spectrally accurate quadrature scheme to evaluate the corresponding layer potentials, so that only a small number of spatial discretization points per particle are required. The resulting discrete sums are computed in $\mathscr{O}(n)$ time, where n denotes the number of particles, using the fast multipole method (FMM). The particle positions and orientations are updated by a high-order time-stepping scheme. We illustrate the accuracy, conditioning and scaling of our solvers with several numerical examples.

97 MATHEMATICS AND COMPUTING↗

Monge-Ampere grids and the multidimensional mapped Fourier method

The efficiency of a numerical method can be greatly improved by combining it with coordinate transformations tailored to a given problem. This is the basis for the mapped Fourier methods. However, obtaining ''good'' coordinate transformations is a major obstacle for this approach in multidimensions. Here, we calculate coordinate transformations based on solving the Monge-Ampere equation. These transformations are combined in the mapped Fourier method and applied to Schroedinger's equation in multidimensions. Dramatic improvements in accuracy compared to the standard Fourier method were observed in eigenvalue calculations for two-dimensional systems. This work indicates that the Monge-Ampere equation may serve as a useful tool for constructing efficient representations for problems in computational quantum mechanics and other fields.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

A Deep Generative Model for Non-Intrusive Identification of EV Charging Profiles

The proliferation of electric vehicles (EVs) brings environmental benefits and technical challenges to power grids. An identification algorithm which can accurately extract individual EV charging profiles out of widely available smart meter measurements has attracted great interests. This paper proposes a non-intrusive identification framework for EV charging profile extraction, which is driven by deep generative models (DGM). First, the proposed DGM is designed as a representation layer embedded into the Markov process and used to model the joint probability distribution of available time-series data. A novel contribution is to approximate posterior distributions by neural networks whose parameters are obtained by variational inference and supervised learning. Second, the EV charging status is inferred from the DGM via dynamic programming. Lastly, the desired EV charging profile can be reconstructed by the rated power of EV models and inferred status. Compared with the benchmark Hidden Markov Models, the proposed framework can better handle noise in data with less computational complexity and better overall accuracy performances with smaller recall. The proposed framework is validated by numerical experiments on the Pecan Street dataset.

33 ADVANCED PROPULSION SYSTEMS↗

Leveraging Multitime Hamilton–Jacobi PDEs for Certain Scientific Machine Learning Problems

Hamilton-Jacobi partial differential equations (HJ PDEs) have deep connections with a wide range of fields, including optimal control, differential games, and imaging sciences. By considering the time variable to be a higher dimensional quantity, HJ PDEs can be extended to the multi-time case. In this paper, we establish a novel theoretical connection between specific optimization problems arising in machine learning and the multi-time Hopf formula, which corresponds to a representation of the solution to certain multi-time HJ PDEs. Through this connection, we increase the interpretability of the training process of certain machine learning applications by showing that when we solve these learning problems, we also solve a multi-time HJ PDE and, by extension, its corresponding optimal control problem. As a first exploration of this connection, we develop the relation between the regularized linear regression problem and the Linear Quadratic Regulator (LQR). We then leverage our theoretical connection to adapt standard LQR solvers (namely, those based on the Riccati ordinary differential equations) to design new training approaches for machine learning. Lastly, we provide some numerical examples that demonstrate the versatility and possible computational advantages of our Riccati-based approach in the context of continual learning, post-training calibration, transfer learning, and sparse dynamics identification.

97 MATHEMATICS AND COMPUTING↗

Quantifying the Impact of Single Bit Flips on Floating Point Arithmetic

In high-end computing, the collective surface area, smaller fabrication sizes, and increasing density of components have led to an increase in the number of observed bit flips. If mechanisms are not in place to detect them, such flips produce silent errors, i.e. the code returns a result that deviates from the desired solution by more than the allowed tolerance and the discrepancy cannot be distinguished from the standard numerical error associated with the algorithm. These phenomena are believed to occur more frequently in DRAM, but logic gates, arithmetic units, and other circuits are also susceptible to bit flips. Previous work has focused on algorithmic techniques for detecting and correcting bit flips in specific data structures, however, they suffer from lack of generality and often times cannot be implemented in heterogeneous computing environment. Our work takes a novel approach to this problem. We focus on quantifying the impact of a single bit flip on specific floating-point operations. We analyze the error induced by flipping specific bits in the most widely used IEEE floating-point representation in an architecture-agnostic manner, i.e., without requiring proprietary information such as bit flip rates and the vendor-specific circuit designs. We initially study dot products of vectors and demonstrate that not all bit flips create a large error and, more importantly, expected value of the relative magnitude of the error is very sensitive on the bit pattern of the binary representation of the exponent, which strongly depends on scaling. Our results are derived analytically and then verified experimentally with Monte Carlo sampling of random vectors. Furthermore, we consider the natural resilience properties of solvers based on the fixed point iteration and we demonstrate how the resilience of the Jacobi method for linear equations can be significantly improved by rescaling the associated matrix.

Elliott, James J.↗

Diagnosing the Nature of Land-Atmosphere Coupling During the 2006-7 Dry/Wet Extremes in the U. S. Southern Great Plains

Land-atmosphere (L-A) interactions play a critical role in determining the diurnal evolution of land surface and planetary boundary layer (PBL) temperature and moisture states and fluxes. In turn, these interactions regulate the strength of the connection between surface moisture and precipitation in a coupled system. To address deficiencies in numerical weather prediction and climate models due to improper treatment of L-A interactions, recent studies have focused on development of diagnostics to quantify the strength and accuracy of the land-PBL coupling at the process-level. In this study, a diagnosis of the nature and impacts of local land-atmosphere coupling (LoCo) during dry and wet extreme conditions is presented using a combination of models and observations during the summers of2006-7 in the U.S. Southern Great Plains. Specifically, the Weather Research and Forecasting (WRF) model has been coupled to NASA's Land Information System (LIS), which provides a flexible and high resolution representation and initialization of land surface physics and states. A range of diagnostics exploring the links and feedbacks between soil moisture and precipitation are examined for the dry/wet regimes of this region, along with the behavior and accuracy of different land-PBL scheme couplings under these conditions. Results demonstrate how LoCo diagnostics can be applied to coupled model components in the context of their integrated impacts on the process-chain connecting the land surface to the PBL and support of hydrological anomalies.

Santanello, Joseph A.↗

Metaplectic geometrical optics for ray-based modeling of caustics: Theory and algorithms

The optimization of radio frequency-wave (RF) systems for fusion experiments is often performed using ray-tracing codes, which rely on the geometrical-optics (GO) approximation. However, GO fails at caustics such as cutoffs and focal points, erroneously predicting the wave intensity to be infinite. This is a critical shortcoming of GO, since the caustic wave intensity is often the quantity of interest, e.g., RF heating. Full-wave modeling can be used instead, but the computational cost limits the speed at which such optimizations can be performed. Here, we have developed a less expensive alternative called metaplectic geometrical optics (MGO). Instead of evolving waves in the usual x (coordinate) or k (spectral) representation, MGO uses a mixed X$\equiv$Ax+Bk representation. By continuously adjusting the matrix coefficients A and B along the rays, one can ensure that GO remains valid in the X coordinates without caustic singularities. The caustic-free result is then mapped back onto the original x space using metaplectic transforms. Here, we overview the MGO theory and review algorithms that will aid the development of an MGO-based ray-tracing code. We show how using orthosymplectic transformations leads to considerable simplifications compared to previously published MGO formulas. We also prove explicitly that MGO exactly reproduces standard GO when evaluated far from caustics (an important property that until now has only been inferred from numerical simulations), and we relate MGO to other semiclassical caustic-removal schemes published in the literature. Finally this discussion is then augmented by an explicit comparison of the computed spectrum for a wave bounded between two cutoffs.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Upgraded viscous flow analysis of multi-element airfoils

A description of an improved version of the NASA/Lockheed multi-element airfoil analysis computer program is presented. The improvements include several major modifications of the aerodynamic model as well as substantial changes of the computer code. The modifications of the aerodynamic model comprise the representation of the boundary layer and wake displacement effects with an equivalent source distribution, the prediction of wake parameters with Green's lag-entrainment method, the calculation of turbulent boundary layer separation with the method of Nash and Hicks, the estimation of the onset of confluent boundary layer separation with a modified form of Goradia's method, and the prediction of profile drag with the formula of Squire and Young. The modifications of the computer program for which the structured approach to computer software development was employed are also described. Important aspects of the structured program development such as the functional decomposition of the aerodynamic theory and its numerical implementation, the analysis of the data flow within the code, and the application of a pseudo code are discussed.

Brune, G. W.↗

Identification of observer/Kalman filter Markov parameters: Theory and experiments

An algorithm to compute Markov parameters of an observer or Kalman filter from experimental input and output data is discussed. The Markov parameters can then be used for identification of a state space representation, with associated Kalman gain or observer gain, for the purpose of controller design. The algorithm is a non-recursive matrix version of two recursive algorithms developed in previous works for different purposes. The relationship between these other algorithms is developed. The new matrix formulation here gives insight into the existence and uniqueness of solutions of certain equations and gives bounds on the proper choice of observer order. It is shown that if one uses data containing noise, and seeks the fastest possible deterministic observer, the deadbeat observer, one instead obtains the Kalman filter, which is the fastest possible observer in the stochastic environment. Results are demonstrated in numerical studies and in experiments on an ten-bay truss structure.

Juang, Jer-Nan↗