Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “multiple iterations”

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 343 records · Page 19
All sourcesNASA NTRSDOE OSTI
NASA NTRS1993

On the synthesis of resonance lines in dynamical models of structured hot-star winds

We examine basic issues involved in synthesizing resonance-line profiles from 1-D, dynamical models of highly structured hot-star winds. Although these models exhibit extensive variations in density as well as velocity, the density scale length is still typically much greater than the Sobolev length. The line transfer is thus treated using a Sobolev approach, as generalized by Rybicki & Hummer (1978) to take proper account of the multiple Sobolev resonances arising from the nonmonotonic velocity field. The resulting reduced-lambda-matrix equation describing nonlocal coupling of the source function is solved by iteration, and line profiles are then derived from formal solution integration using this source function. Two more approximate methods that instead use either a stationary or a structured, local source function yield qualitatively similar line-profiles, but are found to violate photon conservation by 10% or more. The full results suggest that such models may indeed be able to reproduce naturally some of the qualitative properties long noted in observed UV line profiles, such as discrete absorption components in unsaturated lines, or the blue-edge variability in saturated lines. However, these particular models do not yet produce the black absorption troughs commonly observed in saturated lines, and it seems that this and other important discrepancies (e.g., in acceleration time scale of absorption components) may require development of more complete models that include rotation and other 2-D and/or 3-D effects.

Puls, J.↗
DOE OSTI2020

Scalable Deep-Learning-Accelerated Topology Optimization for Additively Manufactured Materials

Topology optimization (TO) is a popular and powerful computational approach for designing novel structures, materials, and devices. Two computational challenges have limited the applicability of TO to a variety of industrial applications. First, a TO problem often involves a large number of design variables to guarantee sufficient expressive power. Second, many TO problems require a large number of expensive physical model simulations, and those simulations cannot be parallelized. To address these issues, we propose a general scalable deep-learning (DL) based TO framework, referred to as SDL-TO, which utilizes parallel schemes in high performance computing (HPC) to accelerate the TO process for designing additively manufactured (AM) materials. Unlike the existing studies of DL for TO, our framework accelerates TO by learning the iterative history data and simultaneously training on the mapping between the given design and its gradient. The surrogate gradient is learned by utilizing parallel computing on multiple CPUs incorporated with a distributed DL training on multiple GPUs. The learned TO gradient enables a fast online update scheme instead of an expensive update based on the physical simulator or solver. Using a local sampling strategy, we achieve to reduce the intrinsic high dimensionality of the design space and improve the training accuracy and the scalability of the SDL-TO framework. The method is demonstrated by benchmark examples and AM materials design for heat conduction. The proposed SDL-TO framework shows competitive performance compared to the baseline methods but significantly reduces the computational cost by a speed up of around 8.6x over the standard TO implementation.

Bi, Sirui↗
DOE OSTI2022

Extension of the energetic particle transport kick model in TRANSP to multiple fast ion species

Abstract Alfvénic instabilities (AEs) are well known to cause enhanced transport of energetic particles (EPs) in fusion devices. Most studies until now have focused on characterizing and understanding AE stability in single-species plasmas heated by neutral beams (NB), where deuterium is typically used as both main plasma species and NB fuel. As the fusion community moves toward fusion reactors that target burning plasma conditions, such as ITER, the single-species picture breaks down. Burning plasmas, which will use a mix of deuterium and tritium (DT) as main fuel, also feature the presence of several supra-thermal fusion products such as alpha particles, protons, helium isotopes and high-energy tritium ions. This work presents the extension of the EP transport kick model implemented in the TRANSP time-dependent tokamak transport code to study the combined effect of multiple EP species on AE stability and, in turn, the response of different EP species to plasma instabilities in terms of their redistribution and losses. Further validation of the enhanced model is planned based on experimental results expected from the JET DT campaign scheduled for 2021, in preparation for ITER plasmas and beyond.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2020

After the hurricane: Validating a resilience assessment methodology

With increasing utility grid outages in the United States, there is growing interest in assessing risk and developing mitigation strategies to reduce the impact of grid outages. Working with the U.S. Air Force, the U.S. Department of Energy's National Renewable Energy Laboratory (NREL) developed a replicable energy resilience assessment methodology and investment decision tool to: (1) identify and score hazards and vulnerabilities at the site level; (2) analyze risks to energy infrastructure; and (3) identify and prioritize energy resilience investments. This work improves on existing resilience assessment methodologies and tools by combining a bottom-up, all-hazards assessment methodology with top-down geographic information system mapping capabilities to provide an innovative, dynamic tool for identifying and prioritizing actionable solutions. This process combines probabilistic forecasting with an iterative approach for continuously updating and reassessing risks to address temporal dynamism. Relationships among systems are modeled and visualized to estimate the effectiveness of resilience actions across multiple interdependent systems and inform financial priorities through cost-difficulty-impact trade-offs. The approach is validated in a case study at Tyndall Air Force Base (AFB) in Florida, which experienced a Category 5 hurricane in 2018. The risks and mitigation strategies identified pre-hurricane are compared with post-hurricane, realized impacts. The assessment effectively identifies risks and actions to increase site energy resilience, but the methodology can be enhanced though greater consideration of the interdependencies between the energy system and related systems like transportation, communication, and food/water systems, which impact the recovery of the energy system and the base.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2026

Plan It Together: Optimizing Across Generation, Transmission, Distribution, and Distributed Energy Resources

Integrated planning holds the promise of unlocking lower total system cost solutions by connecting previously siloed planning processes. This article summarizes the state of bulk and local grid planning today and details multiple new analytical approaches that can enable more holistic planning methods to develop comprehensive solutions to generation, transmission, distribution, and distributed energy resource needs. These include both iterative approaches as well as cooptimization techniques. In addition to the benefits of these methods, the technical and institutional challenges and associated solutions are also discussed.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2026

EvoDiffMol: evolutionary diffusion framework for 3D molecular design with optimized properties

Designing molecules with specific target properties remains a fundamental challenge in computational chemistry. While existing approaches show promise, most rely on simplified representations like SMILES strings or 2D graphs that lack essential three-dimensional geometric information. We present EvoDiffMol, a computational framework that integrates evolutionary algorithms with three-dimensional diffusion models for property-driven molecular generation. The method operates through adaptive evolutionary optimization, where population-based selection guides the generation process toward desired property landscapes. EvoDiffMol supports both unconstrained molecular design and scaffold-constrained generation that preserves fixed substructures while optimizing complementary regions. Comprehensive evaluation demonstrates exceptional performance, achieving the highest drug-likeness score (0.94) among all compared state-of-the-art methods while maintaining excellent validity, uniqueness, and novelty. Beyond single property optimization, the framework demonstrates flexible multi-property optimization capabilities, simultaneously controlling multiple molecular descriptors including synthetic accessibility, lipophilicity, topological polar surface area, and clinically relevant ADMET properties such as cardiotoxicity (hERG) and intestinal permeability (Caco-2). This adaptability spans from simple descriptors to practical pharmaceutical endpoints without requiring complete model retraining. The framework achieves precise control over target property values, generating molecules with properties closely matching specified targets for both single and multiple descriptors. Scaffold-constrained experiments preserve fixed molecular cores while maintaining effective property optimization. The three-dimensional representation offers advantages in maintaining structural validity during iterative optimization, with potential for geometry-aware applications in materials science and drug discovery.

3D molecular generation↗
NASA NTRS1979

A three-dimensional viscous/potential flow interaction analysis method for multi-element wings: Modifications to the potential flow code to allow part-span, high-lift devices and close-interference calculations

The description of the modified code includes details of a doublet subpanel technique in which panels that are close to a velocity calculation point are replaced by a subpanel set. This treatment gives the effect of a higher panel density without increasing the number of unknowns. In particular, the technique removes the close approach problem of the earlier singularity model in which distortions occur in the detailed pressure calculation near panel corners. Removal of this problem allowed a complete wake relaxation and roll-up iterative procedure to be installed in the code. The geometry package developed for the new technique and also for the more general configurations is based on a multiple patch scheme. Each patch has a regular array of panels, but arbitrary relationships are allowed between neighboring panels at the edges of adjacent patches. This provides great versatility for treating general configurations.

Maskew, B.↗
NASA NTRSDate not supplied

Microbial Optical Data Processing: A Key Step in the Metabolic Assessment of Lunar Explorer Instrument for Space Biology Applications (LEIA) and Biosentinel’s Payload Data

The BioSensor payload platform on BioSentinel and LEIA autonomously collects optical data from microbial model organisms in liquid culture. The BioSensor is designed to monitor metabolic activity using absorbance measurements of cell density and alamarBlue, a readily available colorimetric redox indicator dye. BioSentinel, a pioneering NASA CubeSat, uses yeast to study deep space radiation. LEIA investigates radiation and lunar gravity response. The experimental setup includes 16 wells equipped with three LEDs (570, 630, and 850 nm) and their corresponding photodetectors. One well is a calibration control without biology while the rest have desiccated cultures. Autonomous rehydration initiates the experiment. Data from the BioSensor are received from the flight and ground units, enabling comparison to uncover location-based metabolic rate variations. This study presents a Python Jupyter notebook developed for efficient data processing of multiple CSV files containing date and time columns, temperature, and well illumination data. It offers a user-friendly interface while maintaining computational power, automatically recognizing and iteratively processing data files in a user-input path. A Hampel filter with a short window eliminates outlier artifacts from sensor dropout. Because absorbance is a relative measurement, conversion from raw illumination requires defining a “blank” value, so the first data points are averaged to provide the necessary denominator. A cube-root function correction mitigates undesired drift caused by air pockets during the fluidic card filling phase, maintaining optical path length consistency. Beer-Lambert's law is applied to further convert absorbance values to cell and dye form concentrations, the desired science parameters. The processed data are saved and visualized as SVG plots. Future plans include extracting specific science parameters from the processed data like growth rate and metabolic rate, and identification of features corresponding to metabolic and phenotypic shifts such as starvation, shifts from aerobic to anaerobic growth, and osmotic stresses.

Space biology↗
DOE OSTI2022

Optimization of Water-Alternating-CO2 Injection Field Operations Using a Machine-Learning-Assisted Workflow

Summary This paper will present a robust workflow to address multiobjective optimization (MOO) of carbon dioxide (CO2)-enhanced oil recovery (EOR)-sequestration projects with a large number of operational control parameters. Farnsworth unit (FWU) field, a mature oil reservoir undergoing CO2 alternating water injection (CO2-WAG) EOR, will be used as a field case to validate the proposed optimization protocol. The expected outcome of this work would be a repository of Pareto-optimal solutions of multiple objective functions, including oil recovery, carbon storage volume, and project economics. FWU’s numerical model is used to demonstrate the proposed optimization workflow. Because using MOO requires computationally intensive procedures, machine-learning-based proxies are introduced to substitute for the high-fidelity model, thus reducing the total computation overhead. The vector machine regression combined with the Gaussian kernel (Gaussian-SVR) is used to construct proxies. An iterative self-adjusting process prepares the training knowledge base to develop robust proxies and minimizes computational time. The proxies’ hyperparameters will be optimally designed using Bayesian optimization to achieve better generalization performance. Trained proxies will be coupled with multiobjective particle swarm Optimization (MOPSO) protocol to construct the Pareto-front solution repository. The outcomes of this workflow will be a repository containing Pareto-optimal solutions of multiple objectives considered in the CO2-WAG project. The proposed optimization workflow will be compared with another established methodology using a multilayer neural network (MLNN) to validate its feasibility in handling MOO with a large number of parameters to control. Optimization parameters used include operational variables that might be used to control the CO2-WAG process, such as the duration of the water/gas injection period, producer bottomhole pressure (BHP) control, and water injection rate of each well included in the numerical model. It is proved that the workflow coupling Gaussian-SVR proxies and the iterative self-adjusting protocol is more computationally efficient. The MOO process is made more rapid by squeezing the size of the required training knowledge base while maintaining the high accuracy of the optimized results. The outcomes of the optimization study show promising results in successfully establishing the solution repository considering multiple objective functions. Results are also verified by validating the Pareto fronts with simulation results using obtained optimized control parameters. The outcome from this work could provide field operators an opportunity to design a CO2-WAG project using as many inputs as possible from the reservoir models. The proposed work introduces a novel concept that couples Gaussian-SVR proxies with a self-adjusting protocol to increase the computational efficiency of the proposed workflow and to guarantee the high accuracy of the obtained optimized results. More importantly, the workflow can optimize a large number of control parameters used in a complex CO2-WAG process, which greatly extends its utility in solving large-scale MOO problems in various projects with similar desired outcomes.

Energy & Fuels↗
NASA NTRSDate not supplied

Rapid Development of Instrument Thermal Models: Perspectives and Guidelines from NASA Goddard’s Instrument Design Laboratory

The Instrument Design Laboratory (IDL), part of NASA Goddard Space Flight Center’s Integrated Design Center (IDC), is a concurrent and collaborative environment which allows for rapid development of science instrumentation concepts within the span of less than two weeks. Science goals set by a Principal Investigator from government, industry or academia are translated into engineering requirements, from which a team of engineers spanning multiple disciplines use an established study process and a suite of analysis tools to work towards an instrument point design. As part of this process, a staff thermal engineer is tasked with designing a thermal control system which meets all incoming thermal requirements, while iterating real-time with other subsystems to ensure compatibility and functionality as a completed system. Thermal engineers on spaceflight projects typically have weeks or months to develop thermal models. However, the severe time limitation in this conceptual study setting makes thermal design particularly difficult, as rapid thermal modeling solely over the span of a few days is required to develop the instrument thermal design and understand the performance over its intended mission, especially if the instrument concept contains multiple thermal challenges such as dynamic environments or high heat dissipating components. In this paper, the authors provide a condensed guide for the most efficient ways to develop thermal models and conduct thermal analysis within the span of one-to-two weeks, as informed by decades of design experience and best practices in the IDL. The authors also focus on quick methods for determining worst-case thermal environments, deciding which modeling details are essential at this early phase, and quantifying the engineering resources necessary for thermal control. This paper concludes with specific thermal design tips for different instrument types across the electromagnetic spectrum.

Kan Yang↗
DOE OSTI2024

Analytic amplitudes for a pair of Higgs bosons in association with three partons

The pair production of Higgs bosons at the LHC can give information about the triple Higgs boson coupling. We perform an analytic one-loop calculation of the amplitudes for a pair of Higgs bosons in association with three partons, retaining the exact dependence on the quark mass circulating in the loop. These amplitudes constitute the real radiation corrections in the calculation of Higgs boson pair production at next-to-leading order in the strong coupling. The results of an analytic generalised-unitarity computation are simplified via analytic reconstruction in spinor variables. Compact ansätze for kinematic pole residues are iteratively fitted via p-adic evaluations near said poles and subtracted until no pole remains. A new ansatz construction is introduced to minimally parametrise coefficients of amplitudes with multiple massive external legs. The simplified expressions are faster to evaluate than automatic codes and can lead to more stable results near singular regions.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2020

Layer-Parallel Training of Deep Residual Neural Networks

Residual neural networks (ResNets) are a promising class of deep neural networks that have shown excellent performance for a number of learning tasks, e.g., image classification and recognition. Mathematically, ResNet architectures can be interpreted as forward Euler discretizations of a nonlinear initial value problem whose time-dependent control variables represent the weights of the neural network. Hence, training a ResNet can be cast as an optimal control problem of the associated dynamical system. For similar time-dependent optimal control problems arising in engineering applications, parallel-in-time methods have shown notable improvements in scalability. This paper demonstrates the use of those techniques for efficient and effective training of ResNets. The proposed algorithms replace the classical (sequential) forward and backward propagation through the network layers with a parallel nonlinear multigrid iteration applied to the layer domain. This adds a new dimension of parallelism across layers that is attractive when training very deep networks. From this basic idea, we derive multiple layer-parallel methods. The most efficient version employs a simultaneous optimization approach where updates to the network parameters are based on inexact gradient information in order to speed up the training process. Finally, using numerical examples from supervised classification, we demonstrate that the new approach achieves a training performance similar to that of traditional methods, but enables layer-parallelism and thus provides speedup over layer-serial methods through greater concurrency.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Assessing Edge Localized Mode (ELM) Mitigation by Pellet Triggering in Smallest-Possible Affordable Robust Compact (SPARC) Low-Collisionality Discharges

Simulations of pellet edge-localized mode (ELM) triggering with the M3D-C1 code confirmed experimentally observed pellet mass thresholds in low-collisionality DIII-D discharges and predicted extremely large thresholds in ITER [A. Wingen et al, Nucl. Fusion 64, 066015(2024)]. The linear (static) and nonlinear (dynamic) workflows developed in this previous work are now applied to the SPARC tokamak. Multiple synthetic equilibria for the high-performance H-mode scenario are tested for MHD stability without pellets using the linear workflow and ELITE simulations; a stable one is identified. Pellets of various mass are injected along an inboard, near-midplane trajectory in the M3D-C1 simulation. No pellet mass threshold for ELM triggering can be found for realistic pellet masses, concluding that fueling from this injection location would not trigger ELMs. A poloidal scan of possible injection locations shows very high thresholds on the high-field side and reasonable thresholds on the low-field-side (LFS). Based on results of this work the location for a LFS injection was finalized in the SPARC design. For this LFS injection the pellet mass threshold is determined using the linear workflow, aligning well with the poloidal scan.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
NASA NTRS1993

A comparison of multiprocessor scheduling methods for iterative data flow architectures

A comparative study is made between the Algorithm to Architecture Mapping Model (ATAMM) and three other related multiprocessing models from the published literature. The primary focus of all four models is the non-preemptive scheduling of large-grain iterative data flow graphs as required in real-time systems, control applications, signal processing, and pipelined computations. Important characteristics of the models such as injection control, dynamic assignment, multiple node instantiations, static optimum unfolding, range-chart guided scheduling, and mathematical optimization are identified. The models from the literature are compared with the ATAMM for performance, scheduling methods, memory requirements, and complexity of scheduling and design procedures.

Storch, Matthew↗
NASA NTRS1990

Numerical derivative techniques for trajectory optimization

The adoption of robust numerical optimization techniques in trajectory simulation programs has resulted in powerful design and analysis tools. These trajectory simulation/optimization programs are widely used, and a representative list includes the GTS system, the POST program, and newer collocation methods such as OTIS and FONPAC. All of these programs rely on optimization algorithms which require objective function and constraint gradient data during the iteration process. However, most trajectory optimization problems lack simple analytical expressions for these derivatives. In the general case a function evaluation involves integrating aerodynamic, propulsive, and gravity forces over multiple trajectory phases with complex control models. With the newer collocation methods, the integration is replaced by defect constraints and cubic approximations for the state. While analytic gradient expressions can sometimes be derived for trajectory optimization problems, the derivation is cumbersome, time consuming, and prone to mistakes. Fortunately, an alternate method exists for the gradient evaluation, namely finite difference approximations. In this paper some finite difference gradient techniques developed for use with the GTS system are presented. These techniques include methods for computing first and second partial derivatives of single and multiple sets of functions. A key feature of these methods is an error control mechanism which automatically adjusts the perturbation size to obtain accurate derivative values.

Hallman, Wayne P.↗
NASA NTRS2011

Hyperspectral Remote Sensing of Atmosphere and Surface Properties

Atmospheric Infrared Sounder (AIRS), Infrared Atmospheric Sounding Interferometer (IASI), and Cross-track Infrared Sounder (CrIS) are all hyper-spectral satellite sensors with thousands of spectral channels. Top of atmospheric radiance spectra measured by these sensors contain high information content on atmospheric, cloud, and surface properties. Exploring high information content contained in these high spectral resolution spectra is a challenging task due to computation e ort involved in modeling thousands of spectral channels. Usually, only very small fractions (4{10 percent) of the available channels are included in physical retrieval systems or numerical weather forecast (NWP) satellite data assimilations. We will describe a method of simultaneously retrieving atmospheric temperature, moisture, cloud, and surface properties using all available spectral channels without sacrificing computational speed. The essence of the method is to convert channel radiance spectra into super-channels by an Empirical Orthogonal Function (EOF) transformation. Because the EOFs are orthogonal to each other, about 100 super-channels are adequate to capture the information content of the radiance spectra. A Principal Component-based Radiative Transfer Model (PCRTM) developed at NASA Langley Research Center is used to calculate both the super-channel magnitudes and derivatives with respect to atmospheric profiles and other properties. There is no need to perform EOF transformations to convert super channels back to spectral space at each iteration step for a one-dimensional variational retrieval or a NWP data assimilation system. The PCRTM forward model is also capable of calculating radiative contributions due to multiple-layer clouds. The multiple scattering effects of the clouds are efficiently parameterized. A physical retrieval algorithm then performs an inversion of atmospheric, cloud, and surface properties in super channel domain directly therefore both reducing the computational need and preserving the information content of the IASI measurements. The inversion algorithm is based on a non-linear Levenberg-Marquardt method with climatology covariance matrices and a priori information as constraints. One advantage of this approach is that it uses all information content from the hyper-spectral data so that the retrieval is less sensitive to instrument noise and eliminates the need for selecting a sub-set of the channels.

Liu, Xu↗
NASA NTRS2018

Deep Neural Network for Precision Multi-band Infrared Image Segmentation

Image segmentation is one of the fundamental steps in computer vision. Separating targets from background clutter with high precision is a challenging operation for both humans and computers. Currently, segmenting objects from IR images is done by tedious manual work. The implementation of a Deep Neural Network (DNN) to perform precision segmentation of multi-band IR video images is presented. A customized pix2pix DNN with multiple layers of generative encoder/decoder and discriminator architecture is used in the IR image segmentation process. Real and synthetic images and ground truths are employed to train the DNN. Iterative training is performed to achieve optimum accuracy of segmentation using a minimal number of training data. Special training images are created to enhance the missing features and to increase the segmentation accuracy of the objects. Retraining strategies are developed to minimize the DNN training time. Single pixel accuracy has been achieved in IR target boundary segmentation using DNNs. The segmentation accuracy between the customized pix2pix DNN and simple thresholding, GraphCut, simple neural network and ResNet models are compared.

Torres, Gil↗