Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “test vector”

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 1,441 records · Page 80
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Knowledge Oriented Graph Unified Transformer (KOGUT) v0.1

KOGUT — Knowledge Oriented Graph Unified Transformer KOGUT implements the Relational Graph Transformer (RelGT) architecture for knowledge graph link prediction in biological domains, with a primary focus on microbial growth media prediction. While the original RelGT (arXiv:2505.10960) targets relational tables, time series, and multi-table databases, KOGUT adapts this architecture for heterogeneous biological knowledge graphs, providing first-in-class AI predictive models for microbial cultivation. Key Adaptations Beyond Original RelGT: - Knowledge Graph Focus: Applied to biological KGs with semantic node types (taxa, chemicals, media, phenotypes, environments) versus generic relational database tables, trained on the KG-Microbe knowledge graph (1.3M entities, 2.9M edges, 24 relation types). - Multimodal Node Encoding: Integrates node labels, categories, descriptions, and synonyms from KG metadata through learned embedding layers—adapting relational column features to graph node attributes with textual semantics. - Extended K-Hop Subgraph Strategy: Optimized neighborhood sampling (3-hop default, configurable up to 200 nodes) tuned for sparse biological networks, building on the original local-global attention framework with biological relation preservation. - Biolink Predicate Preservation: Type-specific transformations for 24 biological edge semantics (occurs_in, consumes, produces, has_phenotype, subclass_of) beyond standard relational foreign keys, enabling multi-relation link prediction. - Inductive Learning Support: Enables zero-shot predictions for novel taxa through feature-based embeddings (temperature, oxygen requirements, gram stain, cell shape), extending the original transductive relational benchmark scope to uncultured microorganisms. CheapSOTA Performance Optimizations (This Distribution): - VQ-EMA Centroid Attention: Vector quantization with exponential moving average for improved global context modeling (+5-10% MRR improvement). - HDF5 Precomputed Data Loading: One-time preprocessing of k-hop subgraphs to eliminate redundant graph traversals (2-5× training speedup). - Distributed Data Parallel Training: Multi-GPU support for scaling to larger knowledge graphs (tested on 4× NVIDIA A100 GPUs at NERSC Perlmutter). - Mixed Precision Training: Automatic mixed precision (AMP) for memory efficiency and faster training. Advantages Over Standard Knowledge Graph Embedding Models: Combines RelGT's proven multi-element tokenization (features, type, hop, structure) with graph-native biological representations, enabling interpretable link prediction across heterogeneous entities that standard embedding models (TransE, RotatE, ComplEx) and table-based transformers cannot directly model. Achieves near-perfect performance on microbial growth media prediction (MRR: 0.9966, Precision@1: 0.9932, Hit@10: 1.0000) while maintaining explainability through attention-based reasoning over biological pathways. Training Data: - KG-Microbe merged knowledge graph: 1,379,337 nodes, 2,960,472 edges - 24 biological relation types including taxonomic hierarchies, metabolic interactions, phenotype associations, and environmental relationships - Primary prediction task: Growth media suitability for microbial taxa (biolink:occurs_in, 50K edges) - Multi-relation capability: Predicts links for any of the 24 relation types, including chemical consumption/production, phenotype associations, and taxonomic classification Citation: Original RelGT Architecture: Dwivedi et al., "Relational Graph Transformer", arXiv:2505.10960, 2025 KOGUT Implementation: Knowledge Oriented Graph Unified Transformer for Microbial Growth Media Prediction Developed at Lawrence Berkeley National Laboratory (LBNL) Trained on NERSC Perlmutter supercomputer

Joachimiak, Marcin [Lawrence Berkeley National Lab↗
DOE OSTI2018

Solid Mechanics Code Verification: Contrasting Classical and Manufactured Reference Solutions

Verification results for Sierra/SM using inexact reference solutions have often exhibited unsatisfactory convergence behavior. With an understanding of the convergence behavior for these types of tests, one can avoid falsely attributing pathologies of the test with incorrectness of the code. Simple theoretical results highlight that for an inexact reference solution two conditions must be met to observe asymptotic convergence. These conditions, and the resulting types of convergence behaviors, are further illustrated with graphical examples depicting the exact, inexact reference, and sequence of numerical solutions as vectors (in a function space). A stress concentration problem is adopted to contrast convergence behaviors when using inexact (classical linear elastic) and exact (manufactured) reference solutions. Convergence is not initially attained with the classical solution. Convergence with the manufactured solution indicates the convergence failure with the classical reference did not result from code error and provides insight on how for this problem asymptotic convergence could be attained with the classical reference solution by modifying the computational models.

42 ENGINEERING↗
DOE OSTI2025

Thermal Radiation Transport with Tensor Trains

We present a novel tensor network algorithm to solve the time-dependent, gray thermal radiation transport equation. The method invokes a tensor train (TT) decomposition for the specific intensity. The efficiency of this approach is dictated by the rank of the decomposition. When the solution is “low rank,” the memory footprint of the specific intensity solution vector may be significantly compressed. The algorithm, following a step-then-truncate approach of a traditional discrete ordinates method, operates directly on the compressed state vector, thereby enabling large speedups for low-rank solutions. To achieve these speedups, we rely on a recently developed rounding approach based on the Gram-SVD. We detail how familiar S N algorithms for (gray) thermal transport can be mapped to this TT framework and present several numerical examples testing both the optically thick and thin regimes. The TT framework finds low-rank structure and supplies up to ≃60× speedups and ≃1000× compressions for problems demanding large angle counts, thereby enabling previously intractable SN calculations and supplying a promising avenue to mitigate ray effects.

79 ASTRONOMY AND ASTROPHYSICS↗
NASA NTRS1973

Effect of rotor design tip seed on aerodynamic performance of a model VTOL lift fan under static and crossflow conditions

Results are presented for a wind tunnel investigation of three single VTOL lift fan stages designed for the same overall total pressure ratio at different rotor tip speeds. The stages were tested in a model lift fan installed in a wing pod. The three stages had essentially the same aerodynamic performance along the operating line. However, differences in stage thrust characteristics were obtained when a variation in back pressure was imposed on the stages by cross-flow effects and thrust-vectoring louvers.

Stockman, N. O.↗
NASA NTRS1981

A multigrid algorithm for steady transonic potential flows around aerofoils using Newton iteration

The application of multigrid relaxation to transonic potential flow calculation was investigated. Conservative potential flows around aerofoils were taken as test problems. The solution algorithm was based on Newton iteration. It was found that the iteration to the circulation has to be kept outside the multigrid algorithm. To obtain meaningful norms of residuals, difference formulas with asymptotic scaling were introduced. Nonlinear instability problems were solved by upwind differencing using mass flux vector splitting instead of artificial viscosity or artificial density. It is suggested that the algorithms efficiency be increased by improving the iteration on the shock positions even though this is a highly nonlinear process.

Boestoel, J. W.↗
NASA NTRS1987

A vector-based failure detection and isolation algorithm for a dual fail-operational redundant strapdown inertial measurement unit

A vector-based failure detection and isolation technique for a skewed array of two degree-of-freedom inertial sensors is developed. Failure detection is based on comparison of parity equations with a threshold, and isolation is based on comparison of logic variables which are keyed to pass/fail results of the parity test. A multi-level approach to failure detection is used to ensure adequate coverage for the flight control, display, and navigation avionics functions. Sensor error models are introduced to expose the susceptibility of the parity equations to sensor errors and physical separation effects. The algorithm is evaluated in a simulation of a commercial transport operating in a range of light to severe turbulence environments. A bias-jump failure level of 0.2 deg/hr was detected and isolated properly in the light and moderate turbulence environments, but not detected in the extreme turbulence environment. An accelerometer bias-jump failure level of 1.5 milli-g was detected over all turbulence environments. For both types of inertial sensor, hard-over, and null type failures were detected in all environments without incident. The algorithm functioned without false alarm or isolation over all turbulence environments for the runs tested.

Morrell, Frederick R.↗
NASA NTRS1994

Root-sum-square structural strength verification approach

Utilizing a proposed fixture design or some variation thereof, this report presents a verification approach to strength test space flight payload components, electronics boxes, mechanisms, lines, fittings, etc., which traditionally do not lend themselves to classical static loading. The fixture, through use of ordered Euler rotation angles derived herein, can be mounted on existing vibration shakers and can provide an innovative method of applying single axis flight load vectors. The versatile fixture effectively loads protoflight or prototype components in all three axes simultaneously by use of a sinusoidal burst of desired magnitude at less than one-third the first resonant frequency. Cost savings along with improved hardware confidence are shown. The end product is an efficient way to verify experiment hardware for both random vibration and strength.

Lee, Henry M.↗
NASA NTRS2008

Silent Aircraft Initiative Concept Risk Assessment

A risk assessment of the Silent Aircraft Initiative's SAX-40 concept design for extremely low noise has been performed. A NASA team developed a list of 27 risk items, and evaluated the level of risk for each item in terms of the likelihood that the risk would occur and the consequences of the occurrence. The following risk items were identified as high risk, meaning that the combination of likelihood and consequence put them into the top one-fourth of the risk matrix: structures and weight prediction; boundary-layer ingestion (BLI) and inlet design; variable-area exhaust and thrust vectoring; displaced-threshold and continuous descent approach (CDA) operational concepts; cost; human factors; and overall noise performance. Several advanced-technology baseline concepts were created to serve as a basis for comparison to the SAX-40 concept. These comparisons indicate that the SAX-40 would have significantly greater research, development, test, and engineering (RDT&E) and production costs than a conventional aircraft with similar technology levels. Therefore, the cost of obtaining the extremely low noise capability that has been estimated for the SAX-40 is significant. The SAX-40 concept design proved successful in focusing attention toward low noise technologies and in raising public awareness of the issue.

Nickol, Craig L.↗
NASA NTRSDate not supplied

Demonstration of Lidar Sensors for Precision Safe Landing on Planetary Bodies

Missions to solar system bodies must meet increasingly ambitious objectives requiring highly reliable “precision landing”, and “hazard avoidance” capabilities. To meet these needs, we have developed two lidar sensor systems that can be used individually or in concert depending on the mission requirements. One is Navigation Doppler Lidar (NDL) capable of providing vehicle precision vector velocity and ground-relative altitude and the other is an Imaging Flash Lidar for Terrain Relative Navigation and Hazard Detection and Avoidance. Using both lidar sensors together will enable landing anywhere and under any lighting condition. The NDL performance has been extensively characterized through numerous ground and aircraft flight tests and will be soon demonstrated on two lunar landing missions (mid-2023). The Flash Lidar performance is currently being assessed onboard a drone and will soon be tested on helicopter and fixed-wing aircraft platforms. This paper describes both lidar sensors and their expected operation on landing vehicles.

Precision Landing↗
NASA NTRSDate not supplied

Mars Sample Return, Sample Retrieval Lander, Reaction Control System Jet Interaction Supersonic Wind Tunnel Test Overview with CFD Predictions

NASA's Mars Sample Return campaign will be launching several missions over the next decade that will work together to return rock samples from Mars back to Earth. The Sample Retrieval Lander (SRL) will deliver the Mars Ascent Vehicle and fetch rover to the surface of Mars in 2006. Rock samples collected by the Mars 2020 Perseverance rover, landing in early 2021, will be loaded on the the ascent vehicle to be launched into Mars orbit for retrieval by yet another spacecraft. The Sample Retrieval Lander will be a blunt entry capsule similar to past Mars entry vehicles like Mars Science Laboratory and Viking. The vehicle will fly a guided entry, using a small lift vector produced by a non-zero trim angle of attack to eliminate downrange and crossrange position errors at the point of parachute deploy. This energy and heading management is achieved with a reaction control system (RCS) that directs the bank angle of the vehicle and also minimizes unwanted capsule dynamics. The reaction control system and control design is based on the Mars Science Laboratory and Mars 2020 RCS systems. However, due to packaging constraints, the backshell of this new entry vehicle has a different geometry than those earlier designs. To certify the RCS system for flight the project must characterize the jet plume interactions with the capsule backshell that could impair or significantly augment the RCS control authority. This characterization will be done through a combination of computational fluid dynamics (CFD) analysis and wind tunnel test. Two candidate arrangements of the RCS jets have been identified for the SRL vehicle and are currently under evaluation before final selection. The aero/RCS plume interactions of these candidate configurations have been measured in a supersonic wind tunnel test in NASA Langley's Unitary Plan Wind Tunnel. The test was conducted in the fall of 2020 and data is currently being reduced. An overview of the candidate RCS configurations are presented here with an overview of the wind tunnel model design, jet scaling and scaled nozzle design, and the test matrix. Preliminary CFD runs are presented with an assessment of the predicted plumes and their interaction with the wake flow of the vehicle. The predicted effects of the model sting is provided as well. This high fidelity wind tunnel test is being conducted much earlier in the SRL project than would normally be done. The test was funded as part of a CFD evaluation task funded by NASA's Aerosciences Evaluation and Test Capabilities Project. The objective of the evaluation task was to compare the ability of CFD to predict complex flows with data that can be measured in the Langley Unitary Plan Wind Tunnel. RCS Jet interactions were selected as a type of complex flow that is important to NASA missions. In addition to providing useful data to the SRL project, there was added emphasis on quantifying the accuracy of the CFD predictions and wind tunnel test data. An overview of the uncertainty quantification methodologies for computational and experimental portions of this test is presented.

blunt body↗
DOE OSTI2023

Scalable Bayesian optimization with randomized prior networks

Several fundamental problems in science and engineering consist of global optimization tasks involving unknown high-dimensional (black-box) functions that map a set of controllable variables to the outcomes of an expensive experiment. Bayesian Optimization (BO) techniques are known to be effective in tackling global optimization problems using a relatively small number objective function evaluations, but their performance suffers when dealing with high-dimensional outputs. To overcome the major challenge of dimensionality, here we propose a deep learning framework for BO and sequential decision making based on bootstrapped ensembles of neural architectures with randomized priors. Using appropriate architecture choices, we show that the proposed framework can approximate functional relationships between design variables and quantities of interest, even in cases where the latter take values in high-dimensional vector spaces or even infinite-dimensional function spaces. In the context of BO, we augmented the proposed probabilistic surrogates with re-parameterized Monte Carlo approximations of multiple-point (parallel) acquisition functions, as well as methodological extensions for accommodating black-box constraints and multi-fidelity information sources. We test the proposed framework against state-of-the-art methods for BO and demonstrate superior performance across several challenging tasks with high-dimensional outputs, including a constrained multi-fidelity optimization task involving shape optimization of rotor blades in turbo-machinery.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

An arbitrary-order Runge–Kutta discontinuous Galerkin approach to reinitialization for banded conservative level sets

Here, we present a Runge-Kutta discontinuous Galerkin method for solving conservative reinitialization in the context of the conservative level set method. This represents an extension of the method recently proposed by Owkes and Desjardins [21], by solving the level set equations on the refined level set grid and projecting all spatially-dependent variables into the full basis used by the discontinuous Galerkin discretization. By doing so, we achieve the full k+1 order convergence rate in the L1 norm of the level set field predicted for RKDG methods given kth degree basis functions when the level set profile thickness is held constant with grid refinement. Shape and volume errors for the 0.5-contour of the level set, on the other hand, are found to converge between first and second order. We show a variety of test results, including the method of manufactured solutions, reinitialization of a circle and sphere, Zalesak's disk, and deforming columns and spheres, all showing substantial improvements over the high-order finite difference traditional level set method studied for example by Herrmann. We also demonstrate the need for kth order accurate normal vectors, as lower order normals are found to degrade the convergence rate of the method.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2018

Compressed sensing with sparse corruptions: Fault-tolerant sparse collocation approximations

The recovery of approximately sparse or compressible coefficients in a polynomial chaos expansion is a common goal in many modern parametric uncertainty quantification (UQ) problems. However, relatively little effort in UQ has been directed toward theoretical and computational strategies for addressing the sparse corruptions problem, where a small number of measurements are highly corrupted. Such a situation has become pertinent today since modern computational frameworks are sufficiently complex with many interdependent components that may introduce hardware and software failures, some of which can be difficult to detect and result in a highly polluted simulation result. In this paper we present a novel compressive sampling-based theoretical analysis for a regularized t1 minimization algorithm that aims to recover sparse expansion coefficients in the presence of measurement corruptions. Our recovery results are uniform (the theoretical guarantees hold for all compressible signals and compressible corruptions vectors), and prescribe algorithmic regularization parameters in terms of a user-defined a priori estimate on the ratio of measurements that are believed to be corrupted. We also propose an iteratively reweighted optimization algorithm that automatically refines the value of the regularization parameter, and empirically produces superior results. Our numerical results test our framework on several medium-to-high dimensional examples of solutions to parameterized differential equations, and demonstrate the effectiveness of our approach.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1988

Preliminary Orbit Determination System (PODS) for Tracking and Data Relay Satellite System (TDRSS)-tracked target Spacecraft using the homotopy continuation method

The Preliminary Orbit Determination System (PODS) provides early orbit determination capability in the Trajectory Computation and Orbital Products System (TCOPS) for a Tracking and Data Relay Satellite System (TDRSS)-tracked spacecraft. PODS computes a set of orbit states from an a priori estimate and six tracking measurements, consisting of any combination of TDRSS range and Doppler tracking measurements. PODS uses the homotopy continuation method to solve a set of nonlinear equations, and it is particularly effective for the case when the a priori estimate is not well known. Since range and Doppler measurements produce multiple states in PODS, a screening technique selects the desired state. PODS is executed in the TCOPS environment and can directly access all operational data sets. At the completion of the preliminary orbit determination, the PODS-generated state, along with additional tracking measurements, can be directly input to the differential correction (DC) process to generate an improved state. To validate the computational and operational capabilities of PODS, tests were performed using simulated TDRSS tracking measurements for the Cosmic Background Explorer (COBE) satellite and using real TDRSS measurements for the Earth Radiation Budget Satellite (ERBS) and the Solar Mesosphere Explorer (SME) spacecraft. The effects of various measurement combinations, varying arc lengths, and levels of degradation of the a priori state vector on the PODS solutions were considered.

Kirschner, S. M.↗
NASA NTRS1992

Overview of the relevant CFD work at Thiokol Corporation

An in-house developed proprietary advanced computational fluid dynamics code called SHARP (Trademark) is a primary tool for many flow simulations and design analyses. The SHARP code is a time dependent, two dimensional (2-D) axisymmetric numerical solution technique for the compressible Navier-Stokes equations. The solution technique in SHARP uses a vectorizable implicit, second order accurate in time and space, finite volume scheme based on an upwind flux-difference splitting of a Roe-type approximated Riemann solver, Van Leer's flux vector splitting, and a fourth order artificial dissipation scheme with a preconditioning to accelerate the flow solution. Turbulence is simulated by an algebraic model, and ultimately the kappa-epsilon model. Some other capabilities of the code are 2-D two-phase Lagrangian particle tracking and cell blockages. Extensive development and testing has been conducted on the 3-D version of the code with flow, combustion, and turbulence interactions. The emphasis here is on the specific applications of SHARP in Solid Rocket Motor design. Information is given in viewgraph form.

Chwalowski, Pawel↗
NASA NTRS1993

Multi-rate, real time image compression for images dominated by point sources

An image compression system recently developed for compression of digital images dominated by point sources is presented. Encoding consists of minimum-mean removal, vector quantization, adaptive threshold truncation, and modified Huffman encoding. Simulations are presented showing that the peaks corresponding to point sources can be transmitted losslessly for low signal-to-noise ratios (SNR) and high point source densities while maintaining a reduced output bit rate. Encoding and decoding hardware has been built and tested which processes 552,960 12-bit pixels per second at compression rates of 10:1 and 4:1. Simulation results are presented for the 10:1 case only.

Huber, A. Kris↗
NASA NTRS1992

Eigenvalue error analysis of viscously damped structures using a Ritz reduction method

The efficient solution of the eigenvalue problem that results from inserting passive dampers with variable stiffness and damping coefficients into a structure is addressed. Eigenanalysis of reduced models obtained by retaining a number of normal modes augmented with Ritz vectors corresponding to the static solutions resulting from the load patterns introduced by the dampers has been empirically shown to yield excellent approximations to the full eigenvalue problem. An analysis of this technique in the case of a single damper is presented. A priori and a posteriori error estimates are generated and tested on numerical examples. Comparison theorems with modally truncated models and a Markov parameter matching reduced-order model are derived. These theorems corroborate the heuristic that residual flexibility methods improve low-frequency approximation of the system. The analysis leads to other techniques for eigenvalue approximation. Approximate closed-form solutions are derived that include a refinement to eigenvalue derivative methods for approximation. An efficient Newton scheme is also developed. A numerical example is presented demonstrating the effectiveness of each of these methods.

Chu, Cheng-Chih↗
NASA NTRS1992

The wavenumber spectra of scatterometer-derived winds

The accuracy of scatterometer-derived winds using a frequency-domain analysis and simulation is studied. The wavenumber spectra of the Seasat-A-scatterometer (SASS)-derived wind fields have been observed to be accentuated relative to the input wind field. The results of extensive simulations designed to test this observation are reported. Actual SASS measurements of the normalized radar backscatter (NRB) over an orbit (rev) are used as a template to generate simulated NRB measurements. Pointwise estimation of winds from the simulated NRB measurements is accomplished with the ambiguity closest to the true wind selected as the unique wind vector estimate. For comparison, winds are also estimated using a new model-based approach. After wind retrieval, the spectra of the estimated wind fields are computed and compared to the input wind field. The high wavenumber portion of the spectra of the pointwise estimated winds was higher than the spectra of the true winds by an amount which depends on the wind speed variance.

Long, D. G.↗