SEARCH · Engineering Papers

Results for “parameter compression”

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

SourceResearch connectionFrom yearThrough year

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

At least 631 records · Page 35

All sources NASA NTRS DOE OSTI

DOE OSTI2023

Data compression and inference in cosmology with self-supervised machine learning

ABSTRACT The influx of massive amounts of data from current and upcoming cosmological surveys necessitates compression schemes that can efficiently summarize the data with minimal loss of information. We introduce a method that leverages the paradigm of self-supervised machine learning in a novel manner to construct representative summaries of massive data sets using simulation-based augmentations. Deploying the method on hydrodynamical cosmological simulations, we show that it can deliver highly informative summaries, which can be used for a variety of downstream tasks, including precise and accurate parameter inference. We demonstrate how this paradigm can be used to construct summary representations that are insensitive to prescribed systematic effects, such as the influence of baryonic physics. Our results indicate that self-supervised machine learning techniques offer a promising new approach for compression of cosmological data as well as its analysis.

Astronomy & Astrophysics↗

NASA NTRS1984

Video compression using conditional replenishment and motion prediction

A study of a low-rate monochrome video compression system is presentd in this paper. This system is a conditional-replenishment coder that uses two-dimensional Walsh-transform coding within each video frame. The conditional-replenishment algorithm works by transmitting only the portions of an image that are changing in time. This system is augmented with a motn-prediction algorithm that measures spatial dispalcement parameters from frame to frame, and codes the data using these parameters. A comparison is made between the conditional-replenishment system with, and without, the motion-predictinalgorthm. Subsampling in time is ued to maintain the data rate rate at a fixed value. Average bit rates of 1 bit/picture element (pel) to 1/16 bit/pel are considered. The resultant performance of the compression simulations is presented in terms of the average frame rates produced.

Hein, D. N.↗

DOE OSTI2022

Efficient and universal characterization of atomic structures through a topological graph order parameter

Abstract A graph-based order parameter, based on the topology of the graph itself, is introduced for the characterization of atomistic structures. The order parameter is universal to any material/chemical system and is transferable to all structural geometries. Four sets of data are used to validate both the generalizability and accuracy of the algorithm: (1) liquid lithium configurations spanning up to 300 GPa, (2) condensed phases of carbon along with nanotubes and buckyballs at ambient and high temperature, (3) a diverse set of aluminum configurations including surfaces, compressed and expanded lattices, point defects, grain boundaries, liquids, nanoparticles, all at nonzero temperatures, and (4) eleven niobium oxide crystal phases generated with ab initio molecular dynamics. We compare our proposed method to existing, state-of-the-art methods for the cases of aluminum and niobium oxide. Our order parameter uniquely classifies every configuration and outperforms all studied existing methods, opening the door for its use in a multitude of complex application spaces that can require fine structure-level characterization of atomistic graphs.

36 MATERIALS SCIENCE↗

DOE OSTI2023

Visualization and efficient generation of constrained high-dimensional theoretical parameter spaces

We describe a set of novel methods for efficiently sampling high-dimensional parameter spaces of physical theories defined at high energies, but constrained by experimental measurements made at lower energies. Often, theoretical models such as supersymmetry are defined by many parameters, $\mathcal{O}$(10 – 100), expressed at high energies, while relevant experimental constraints are often defined at much lower energies, preventing them from directly ruling out portions of the space. Instead, the low-energy constraints define a complex, potentially non-contiguous subspace of the theory parameters. Naive scanning of the theory space for points which satisfy the low-energy constraints is hopelessly inefficient due to the high dimensionality, and the inverse problem is considered intractable. As a result, many theoretical spaces remain under-explored. We introduce a class of modified generative autoencoders, which attack this problem by mapping the high-dimensional parameter space to a structured low-dimensional latent space, allowing for easy visualization and efficient generation of theory points which satisfy experimental constraints. An extension without dimensional compression, which focuses on limiting potential information loss, is also introduced.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

DOE OSTI2020

Fast strength model characterization using Bayesian statistics

A variety of flow stress models exist with new models constantly being developed. These models aim to approximate the strength of materials in a variety of regimes from quasistatic loading through shock scenarios. All models contain an array of parameters which need to be tuned to the material under study. Some models perform well under limited conditions, requiring adjustment of the parameters when venturing outside of those predefined ranges. Other models perform well over a wide range of conditions with a set of parameters, but may be outperformed by other models optimized on a tighter range of conditions. Recent research by Los Alamos demonstrated the ability to optimize the Johnson Cook (JC) model using a set of 3 plate-impact experiments on Aluminum. They utilized Bayesian statistics and emulation to determine optimal parameters for the model with a quantification of parameter uncertainty. We present an extension of this capability to incorporate velocimetry from plate-impact tests, stress-strain data from split Hopkinson pressure bar and quasistatic compression tests, plus profiles from Taylor cylinders in a unified fashion. Statistically robust comparisons of the performance and uncertainty of different realizations of the JC flow stress model were carried out based on calibration to several possible combinations of these three different experiment types.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

DOE OSTI2021

A comparison of Raman and pyrometry dynamic temperature measurements of shocked cyclohexane

Optical pyrometry is a well-developed, broadly applicable method to determine temperature in dynamic compression experiments. However, it measures radiation from only the skin depth of the radiant material, is highly susceptible to spurious backgrounds, and requires an often-unmeasurable free parameter—the sample dynamic emissivity. Raman spectroscopy offers a method to determine temperature directly from fundamental theory, but it is practically limited in applications to a small subset of shock experiments. In this work, we describe an experiment allowing simultaneous application of Raman and pyrometric temperature techniques to benchmark a specific instance of pyrometry. We attempted both measurements on multiple experiments and compare the successful temperature results between the separate but highly repeatable shots. Here, we find that dynamic shock temperatures inferred from Raman spectroscopy and pyrometry generally agree within the uncertainties of each measurement, although those uncertainties are larger than typical for a specially designed pyrometry experiment on a shocked metal sample.

74 ATOMIC AND MOLECULAR PHYSICS↗

DOE OSTI2021

Experimental quantum learning of a spectral decomposition

Currently available quantum hardware allows for small-scale implementations of quantum machine learning algorithms. Such experiments aid the search for applications of quantum computers by benchmarking the near-term feasibility of candidate algorithms. Here we demonstrate the quantum learning of a two-qubit unitary by a sequence of three parameterized quantum circuits containing a total of 21 variational parameters. Moreover, we variationally diagonalize the unitary to learn its spectral decomposition, i.e., its eigenvalues and eigenvectors. We illustrate how this can be used as a subroutine to compress the depth of dynamical quantum simulations. One can view our implementation as a demonstration of entanglement-enhanced machine learning, as only a single (entangled) training data pair is required to learn a 4 × 4 unitary matrix.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

DOE OSTI2023

The impact of physiologically relevant temperatures on physical properties of thylakoid membranes: a molecular dynamics study

Thylakoid membranes are energy-converting membranes with a unique lipid composition. Though the membranes are primarily composed of proteins, their photosynthetic function is strongly influenced by the lipid constituents. Here we characterize, with molecular dynamics (MD) simulations, lipid bilayers with compositions representative of plant thylakoid membranes. We determine, in a wide range of temperatures, the physical parameters of the model membranes which are relevant for the photosynthetic function. We found a marked impact of temperature on membrane permeability due to a combination of increased compressibility and curvature of the membrane at elevated temperatures. With increasing temperatures, we observed increasingly smeared transmembrane density profiles of the membrane forming lipid headgroups predicting increased membrane flexibility. The diffusion coefficient of the lipids increased with temperature without apparent specificity for lipid species. Instead of a comprehensive experimental dataset in the relevant temperature range, we quantitatively compared and validated our MD results with MD simulations on a dipalmitoylphosphatidylcholine model system.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2025

FPGA-accelerated SpeckleNN with SNL for real-time X-ray single-particle imaging

We present the implementation of a specialized version of our previously published unified embedding model, SpeckleNN, for real-time speckle pattern classification in X-ray Single-Particle Imaging (SPI), using the SLAC Neural Network Library (SNL) on an FPGA platform. This hardware realization transitions SpeckleNN from a prototypic model into a practical edge solution, optimized for running inference near the detector in high-throughput X-ray free-electron laser (XFEL) facilities, such as those found at the Linac Coherent Light Source (LCLS). To address the resource constraints inherent in FPGAs, we developed a more specialized version of SpeckleNN. The original model, which was designed for broader classification across multiple biological samples, comprised ~5.6 million parameters. The new implementation, while reducing the parameter count to 64.6K (a 98.8% reduction), focuses on maintaining the model's essential functionality for real-time operation, achieving an accuracy of 90%. Furthermore, we compressed the latent space from 128 to 50 dimensions. This implementation was demonstrated on the KCU1500 FPGA board, utilizing 71% of available DSPs, 75% of LUTs, and 48% of FFs, with an average power consumption of 9.4W according to the Vivado post-implementation report. The FPGA performed inference on a single image with a latency of 45.015 microseconds at a 200 MHz clock rate. In comparison, running the same inference on an NVIDIA A100 GPU resulted in an average power consumption of ~73W and an image processing latency of around 400 microseconds. Our FPGA-accelerated version of SpeckleNN demonstrated significant improvements, achieving an 8.9 × speedup and a 7.8 × reduction in power consumption compared to the GPU implementation. Key advancements include model specialization and dynamic weight loading through SNL, which eliminates the need for time-consuming FPGA design re-synthesis, allowing fast and continuous deployment of models (re)trained online. These innovations enable real-time adaptive classification and efficient vetoing of speckle patterns, making SpeckleNN more suited for deployment in XFEL facilities. This implementation has the potential to significantly accelerate SPI experiments and enhance adaptability to evolving experimental conditions.

47 OTHER INSTRUMENTATION↗

NASA NTRS1973

Nonlinear problems of the theory of heterogeneous slightly curved shells

An account if given of the variational method of the solution of physically and geometrically nonlinear problems of the theory of heterogeneous slightly curved shells. Examined are the bending and supercritical behavior of plates and conical and spherical cupolas of variable thickness in a temperature field, taking into account the dependence of the elastic parameters on temperature. The bending, stability in general and load-bearing capacity of flexible isotropic elastic-plastic shells with different criteria of plasticity, taking into account compressibility and hardening. The effect of the plastic heterogeneity caused by heat treatment, surface work hardening and irradiation by fast neutron flux is investigated. Some problems of the dynamic behavior of flexible shells are solved. Calculations are performed in high approximations. Considerable attention is given to the construction of a machine algorithm and to the checking of the convergence of iterative processes.

Kantor, B. Y.↗

NASA NTRS1974

Calculation of flow distribution in large radius ratio stages of axial flow turbines and comparison of theory and experiment

A method of calculating stage parameters and flow distribution of axial turbines is described. The governing equations apply to space between the blade rows and are based on the assumption of rotationally symmetrical, compressible, adiabatic flow conditions. Results are presented for stage design and flow analysis calculations. Theoretical results from the calculation system are compared with experimental data from low pressure steam turbine tests.

Herzog, J.↗

NASA NTRS1992

LEWICE/E: An Euler based ice accretion code

A new version of the LEWICE ice accretion computer code was developed which calculates the ice growth on two dimensional surfaces, incorporating the effects of compressibility through the solution of the Euler equations. The code is modular and contains separate stand-alone program elements that create a grid, calculate the flow field parameters, calculate the droplet trajectory paths, determine the amount of ice growth, and plot results. This code increases the applicability of ice accretion predictions by allowing calculations at higher Mach numbers. The new elements of the code are described. Calculated results are compared to experiment for several cases, including a LEWICE example case and a thin airfoil section at a Mach number of 0.58.

Potapczuk, Mark G.↗

NASA NTRS1992

LEWICE/E - An Euler based ice accretion code

Potapczuk, Mark G.↗

NASA NTRS1956

The Compressible Laminar Boundary Layer with Heat Transfer and Arbitrary Pressure Gradient

An approximate method for the calculation of the compressible laminar boundary layer with heat transfer and arbitrary pressure gradient, based on Thwaites' correlation concept, is presented. With the definition of dimensionless shear and heat-transfer parameters and an assumed correlation of these parameters in terms of a momentum parameter, a complete system of relations for calculating skin friction and heat transfer results. Knowledge of velocity or temperature profiles is not necessary in using this calculation method. When the method is applied to a convergent-divergent, axially symmetric rocket nozzle, it shows that high rates of heat transfer are obtained at the initial stagnation point and at the throat of the nozzle. Also indicated are negative displacement thicknesses in the convergent portion of the nozzle; these occur because of the high density within the lower portions of the cooled boundary layer. (author)

Cohen, Clarence B↗

NASA NTRS2013

Radar Range Sidelobe Reduction Using Adaptive Pulse Compression Technique

Pulse compression has been widely used in radars so that low-power, long RF pulses can be transmitted, rather than a highpower short pulse. Pulse compression radars offer a number of advantages over high-power short pulsed radars, such as no need of high-power RF circuitry, no need of high-voltage electronics, compact size and light weight, better range resolution, and better reliability. However, range sidelobe associated with pulse compression has prevented the use of this technique on spaceborne radars since surface returns detected by range sidelobes may mask the returns from a nearby weak cloud or precipitation particles. Research on adaptive pulse compression was carried out utilizing a field-programmable gate array (FPGA) waveform generation board and a radar transceiver simulator. The results have shown significant improvements in pulse compression sidelobe performance. Microwave and millimeter-wave radars present many technological challenges for Earth and planetary science applications. The traditional tube-based radars use high-voltage power supply/modulators and high-power RF transmitters; therefore, these radars usually have large size, heavy weight, and reliability issues for space and airborne platforms. Pulse compression technology has provided a path toward meeting many of these radar challenges. Recent advances in digital waveform generation, digital receivers, and solid-state power amplifiers have opened a new era for applying pulse compression to the development of compact and high-performance airborne and spaceborne remote sensing radars. The primary objective of this innovative effort is to develop and test a new pulse compression technique to achieve ultrarange sidelobes so that this technique can be applied to spaceborne, airborne, and ground-based remote sensing radars to meet future science requirements. By using digital waveform generation, digital receiver, and solid-state power amplifier technologies, this improved pulse compression technique could bring significant impact on future radar development. The novel feature of this innovation is the non-linear FM (NLFM) waveform design. The traditional linear FM has the limit (-20 log BT -3 dB) for achieving ultra-low-range sidelobe in pulse compression. For this study, a different combination of 20- or 40-microsecond chirp pulse width and 2- or 4-MHz chirp bandwidth was used. These are typical operational parameters for airborne or spaceborne weather radars. The NLFM waveform design was then implemented on a FPGA board to generate a real chirp signal, which was then sent to the radar transceiver simulator. The final results have shown significant improvement on sidelobe performance compared to that obtained using a traditional linear FM chirp.

Li, Lihua↗

NASA NTRSDate not supplied

Latest Development in Radiative Transfer Models and Retrieval Algorithms Using Principal Components

The radiative transfer model (RTMs) has a wide range of applications in satellite remote sensing and atmospheric radiation applications. However, millions of line-by-line (LBL) radiative transfer calculations at fine monochromatic frequencies are needed in order to properly calculate spectral contributions of water vapor and trace gases in the atmosphere in infrared and solar spectral regions. Therefore, fast and accurate RTMs are needed to efficiently process large amount of satellite data. A Principal Component-based Radiative Transfer Model (PCRTM) was first developed in 2004 at NASA Langley Research Centre to fulfil this need. By using PC-compression, one can reduce the data dimension significantly while maintaining original information content. The PCRTM can directly compute PC-scores and their derivatives with respect to retrieved parameters. The PCRTM can simulate the top-of-atmosphere (TOA) radiance or reflectance spectra from 0.250 µm (400000 cm-1) to 2000 µm (50 cm-1) with several orders of magnitude faster speed as compared to a LBL RTM. It is also extremely accurate compared to LBL RTM benchmarks (0.03 K RMS error in IR and 0.05% in solar). The PCRTM model has been developed for hyperspectral sensors such as AIRS, CrIS, IASI, NAST-I, SHIS, FIRST, and CLARREO-IR in thermal IR spectral region and CLARREO-Solar, CPF, TEMPO, EMIT, OMI, and SCIAMACHY in solar spectral region. The PCRTM accuracy has been demonstrated via RTM intercomparisons and with real satellite observations from AIRS, CrIS, IASI, SCHIAMACHY, and EMIT etc. In this presentation, we will describe two PCRTM-based inversion algorithms to retrieve atmospheric temperature, water vapor, and trace gas profiles, as well as cloud and surface properties from hyperspectral sounders such as AIRS, CrIS, IASI, and NAST-I. The first one is called Single Fieldof-view Sounder Atmospheric Product (SiFSAP) algorithm. It provides L2 products with 9-times higher area spatial resolution as compared to current cloud-clearing sounder algorithms. The SiFSAP L2 and L3 products are available at NASA's Goddard Earth Sciences Data and Information Services Center (GES DISC) for public access. The second inversion algorithm is called Climate Fingerprinting Atmospheric Product (ClimFiSP) algorithm. It is designed to produce high quality climate products (trends, anomalies, daily, and monthly profiles of temperature, water vapor, traces, clouds, and surface properties) from multiple satellite sensors such AIRS on Aqua and CrIS on multiple satellites. This product will be available at NASA GES DISC later this year. The PCRTMbased high fidelity simulators for CLARREO and CPF have been used for sensor performance trade studies, algorithm development, and inter-satellite calibrations. We have also used PCRTM generated TOA radiance spectra to train an AI-based algorithm and successfully retrieved cloud properties from EMIT solar hyperspectral imagers.

PCRTM↗

DOE OSTI2021

Determining failure properties of as-received and hydrided unirradiated Zircaloy-4 from ring compression tests

An integrated experimental and computational approach was developed to determine the plastic strain at fracture and to investigate the fracture energy of as-received and hydrided unirradiated Zircaloy-4 nuclear cladding tubes in the hoop direction at room temperature from ring compression testing (RCT). This work builds on previous methods in the literature that were developed to obtain the mechanical properties of nuclear cladding before fracture (Young’s modulus, yield stress, and strain-hardening parameters) from experimental RCT results and extends the characterization of this material class to include material failure properties. Furthermore, a ductile damage approach was developed and implemented in a validated finite element model to predict material failure, evaluate fracture energy, and quantify the plastic strain at fracture initiation. The hydrogen content of the Zircaloy-4 specimens was varied up to 610 wppm when the failure became brittle. As expected, increasing hydrogen content caused embrittlement of the Zircaloy-4 samples tested and the plastic strain at fracture and fracture energy to decrease.

36 MATERIALS SCIENCE↗

DOE OSTI2021

Effects of radiation damage and precipitate distribution on micro-pillar compression testing of irradiated CuCrZr

The results of small-scale mechanical tests are convoluted by the so-called size effect, whereby materials appear stronger when the scale of the test is reduced to the order of microns or less. The dimensional range over which this occurs has been shown to be linked to a change in sample microstructure, such as the addition of defects induced by irradiation. To investigate this response, a CuCrZr alloy was subjected to proton irradiation and mechanically tested using micro compression of pillars with a range in size. It was found that irradiation defects dominate over the extrinsic size effect and the sensitivity to differences in precipitate microstructure was also somewhat reduced, suggesting that size-independent results could be obtained from much smaller test volumes in irradiated material compared to their non-irradiated counterparts. Lastly, comparison was made between the increase in yield strength predicted by models and the experimentally measured values to establish the key parameters driving the strengthening behaviour.

36 MATERIALS SCIENCE↗