Search NASA⌕ Search

SEARCH · Search NASA

Results for “Convolution”

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 1,513 records · Page 84

A modified time-of-flight method for precise determination of high speed ratios in molecular beams

Time-of-flight (TOF) is a standard experimental technique for determining, among others, the speed ratio S (velocity spread) of a molecular beam. The speed ratio is a measure for the monochromaticity of the beam and an accurate determination of S is crucial for various applications, for example, for characterising chromatic aberrations in focussing experiments related to helium microscopy or for precise measurements of surface phonons and surface structures in molecular beam scattering experiments. For both of these applications, it is desirable to have as high a speed ratio as possible. Molecular beam TOF measurements are typically performed by chopping the beam using a rotating chopper with one or more slit openings. The TOF spectra are evaluated using a standard deconvolution method. However, for higher speed ratios, this method is very sensitive to errors related to the determination of the slit width and the beam diameter. The exact sensitivity depends on the beam diameter, the number of slits, the chopper radius, and the chopper rotation frequency. We present a modified method suitable for the evaluation of TOF measurements of high speed ratio beams. The modified method is based on a systematic variation of the chopper convolution parameters so that a set of independent measurements that can be fitted with an appropriate function are obtained. We show that with this modified method, it is possible to reduce the error by typically one order of magnitude compared to the standard method.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Design of a portable optical emission tomography system for microwave induced compact plasma for visible to near-infrared emission lines

A new non-invasive diagnostic system is developed for Microwave Induced Plasma (MIP) to reconstruct tomographic images of a 2D emission profile. A compact MIP system has wide application in industry as well as research application such as thrusters for space propulsion, high current ion beams, and creation of negative ions for heating of fusion plasma. Emission profile depends on two crucial parameters, namely, the electron temperature and density (over the entire spatial extent) of the plasma system. Emission tomography provides basic understanding of plasmas and it is very useful to monitor internal structure of plasma phenomena without disturbing its actual processes. This paper presents development of a compact, modular, and versatile Optical Emission Tomography (OET) tool for a cylindrical, magnetically confined MIP system. It has eight slit-hole cameras and each consisting of a complementary metal–oxide–semiconductor linear image sensor for light detection. The optical noise is reduced by using aspheric lens and interference band-pass filters in each camera. The entire cylindrical plasma can be scanned with automated sliding ring mechanism arranged in fan-beam data collection geometry. The design of the camera includes a unique possibility to incorporate different filters to get the particular wavelength light from the plasma. This OET system includes selected band-pass filters for particular argon emission 750 nm, 772 nm, and 811 nm lines and hydrogen emission H{sub α} (656 nm) and H{sub β} (486 nm) lines. Convolution back projection algorithm is used to obtain the tomographic images of plasma emission line. The paper mainly focuses on (a) design of OET system in detail and (b) study of emission profile for 750 nm argon emission lines to validate the system design.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Double crystal monochromator controlled by integrated computing on BL07A in New SUBARU, Japan

The BL07A beamline in New SUBARU, University of Hyogo, has been used for many studies of new materials. A new double crystal monochromator controlled by integrated computing was designed and installed in the beamline in 2014. In this report we will discuss the unique features of this new monochromator, MKZ-7NS. This monochromator was not designed exclusively for use in BL07A; on the contrary, it was designed to be installed at low cost in various beamlines to facilitate the industrial applications of medium-scale synchrotron radiation facilities. Thus, the design of the monochromator utilized common packages that can satisfy the wide variety of specifications required at different synchrotron radiation facilities. This monochromator can be easily optimized for any beamline due to the fact that a few control parameters can be suitably customized. The beam offset can be fixed precisely even if one of the two slave axes is omitted. This design reduces the convolution of mechanical errors. Moreover, the monochromator’s control mechanism is very compact, making it possible to reduce the size of the vacuum chamber can be made smaller.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

A new data processing technique for Rayleigh-Taylor instability growth experiments

Typical face-on experiments for Rayleigh-Taylor instability study involve the time-resolved radiography of an accelerated foil with line-of-sight of the radiography along the direction of motion. The usual method which derives perturbation amplitudes from the face-on images reverses the actual image transmission procedure, so the obtained results will have a large error in the case of large optical depth. In order to improve the accuracy of data processing, a new data processing technique has been developed to process the face-on images. This technique based on convolution theorem, refined solutions of optical depth can be achieved by solving equations. Furthermore, we discuss both techniques for image processing, including the influence of modulation transfer function of imaging system and the backlighter spatial profile. Besides, we use the two methods to the process the experimental results in Shenguang-II laser facility and the comparison shows that the new method effectively improve the accuracy of data processing.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Dynamic measurements and simulations of airborne picolitre-droplet coalescence in holographic optical tweezers

We report studies of the coalescence of pairs of picolitre aerosol droplets manipulated with holographic optical tweezers, probing the shape relaxation dynamics following coalescence by simultaneously monitoring the intensity of elastic backscattered light (EBL) from the trapping laser beam (time resolution on the order of 100 ns) while recording high frame rate camera images (time resolution <10 μs). The goals of this work are to: resolve the dynamics of droplet coalescence in holographic optical traps; assign the origin of key features in the time-dependent EBL intensity; and validate the use of the EBL alone to precisely determine droplet surface tension and viscosity. For low viscosity droplets, two sequential processes are evident: binary coalescence first results from the overlap of the optical traps on the time scale of microseconds followed by the recapture of the composite droplet in an optical trap on the time scale of milliseconds. As droplet viscosity increases, the relaxation in droplet shape eventually occurs on the same time scale as recapture, resulting in a convoluted evolution of the EBL intensity that inhibits quantitative determination of the relaxation time scale. Droplet coalescence was simulated using a computational framework to validate both experimental approaches. The results indicate that time-dependent monitoring of droplet shape from the EBL intensity allows for robust determination of properties such as surface tension and viscosity. Finally, the potential of high frame rate imaging to examine the coalescence of dissimilar viscosity droplets is discussed.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

The separated electric and magnetic field responses of luminescent bacteria exposed to pulsed microwave irradiation

Electromagnetic fields (EMFs) are ubiquitous in the digital world we inhabit, with microwave and millimetre wave sources of non-ionizing radiation employed extensively in electronics and communications, e.g., in mobile phones and Wi-Fi. Indeed, the advent of 5G systems and the “internet of things” is likely to lead to massive densification of wireless networks. Whilst the thermal effects of EMFs on biological systems are well characterised, their putative non-thermal effects remain a controversial subject. Here, we use the bioluminescent marine bacterium, Vibrio fischeri, to monitor the effects of pulsed microwave electromagnetic fields, of nominal frequency 2.5 GHz, on light emission. Separated electric and magnetic field effects were investigated using a resonant microwave cavity, within which the maxima of each field are separated. For pulsed electric field exposure, the bacteria gave reproducible responses and recovery in light emission. At the lowest pulsed duty cycle (1.25%) and after short durations (100 ms) of exposure to the electric field at power levels of 4.5 W rms, we observed an initial stimulation of bioluminescence, whereas successive microwave pulses became inhibitory. Much of this behaviour is due to thermal effects, as the bacterial light output is very sensitive to the local temperature. Conversely, magnetic field exposure gave no measurable short-term responses even at the highest power levels of 32 W rms. Thus, we were able to detect, de-convolute, and evaluate independently the effects of separated electric and magnetic fields on exposure of a luminescent biological system to microwave irradiation.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Structural refinement of Pbnm-type perovskite films from analysis of half-order diffraction peaks

Engineering structural modifications of epitaxial perovskite thin films is an effective route to induce new functionalities or enhance existing properties due to the close relation of the electronic ground state to the local bonding environment. As such, there is a necessity to systematically refine and precisely quantify these structural displacements, particularly those of the oxygen octahedra, which is a challenge due to the weak scattering factor of oxygen and the small diffraction volume of thin films. In this paper, we present an optimized algorithm to refine the octahedral rotation angles using specific unit-cell-doubling half-order diffraction peaks for the a - a - c + Pbnm structure. The oxygen and A-site positions can be obtained by minimizing the squared-error between calculated and experimentally determined peak intensities using the (1/2 1/2 3/2) and (1/2 1/2 5/2) reflections to determine the rotation angle α about in-plane axes and the (1/2 5/2 1), (1/2 3/2 1), and (1/2 3/2 2) reflections to determine the rotation angle γ about the out-of-plane axis, whereas the convoluting A-site displacements associated with the octahedral rotation pattern can be determined using (1 1 1/2) and (1/2 1/2 1/2) reflections to independently determine A-site positions. The validity of the approach is confirmed by applying the refinement procedure to determine the A-site and oxygen displacements in a NdGaO 3 single crystal. Finally, the ability to refine both the oxygen and A-site displacements relative to the undistorted perovskite structure enables a deeper understanding of how structural modifications alter functionality properties in epitaxial films exhibiting this commonly occurring crystal structure.

74 ATOMIC AND MOLECULAR PHYSICS↗

Utilization of outer-midplane collector probes with isotopically enriched tungsten tracer particles for impurity transport studies in the scrape-off layer of DIII-D (invited)

Triplet sets of replaceable graphite rod collector probes (CPs), each with collection surfaces on opposing faces and oriented normal to the magnetic field, were inserted at the outboard mid-plane of DIII-D to study divertor tungsten (W) transport in the Scrape-Off Layer (SOL). Each CP collects particles along field lines with different parallel sampling lengths (determined by the rod diameters and SOL transport) giving radial profiles from the main wall inward to R-R sep ~ 6 cm. The CPs were deployed in a first-of-a-kind experiment using two toroidal rings of distinguishable isotopically enriched, W-coated divertor tiles installed at 2 poloidal locations in the divertor. Post-mortem Rutherford backscatter spectrometry of the surface of the CPs provided areal density profiles of elemental W coverage. Higher W content was measured on the probe side facing along the field lines toward the inner target indicating higher concentration of W in the plasma upstream of the CP, even though the W-coated rings were in the outer target region of the divertor. Inductively coupled plasma mass spectroscopy validates the isotopic tracer technique through analysis of CPs exposed during L-mode discharges with the outer strike point on the isotopically enriched W coated-tile ring. The contribution from each divertor ring of W to the deposition profiles found on the mid-plane collector probes was able to be de-convoluted using a stable isotope mixing model. Furthermore the results provided quantitative information on the W source and transport from specific poloidal locations within the lower divertor region.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

On the memory kernel and the reduced system propagator

We relate the memory kernel in the Nakajima–Zwanzig–Mori time-convolution approach to the reduced system propagator which is often used to obtain the kernel in the Tokuyama–Mori time-convolutionless approach. The connection provides a robust and simple formalism to compute the memory kernel for a generalized system-bath model circumventing the need to compute high order system-bath observables, thus streamlining the use of numerically exact solvers for calculating the memory kernel. We illustrate this for a model system with electron-electron and electron-phonon couplings, driven away from equilibrium.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Prediction of atomization energy using graph kernel and active learning

Data-driven prediction of molecular properties presents unique challenges to the design of machine learning methods concerning data structure/dimensionality, symmetry adaption, and confidence management. In this paper, we present a kernel-based pipeline that can learn and predict the atomization energy of molecules with high accuracy. The framework employs Gaussian process regression to perform predictions based on the similarity between molecules, which is computed using the marginalized graph kernel. To apply the marginalized graph kernel, a spatial adjacency rule is first employed to convert molecules into graphs whose vertices and edges are labeled by elements and interatomic distances, respectively. We then derive formulas for the efficient evaluation of the kernel. Specific functional components for the marginalized graph kernel are proposed, while the effects of the associated hyperparameters on accuracy and predictive confidence are examined. We show that the graph kernel is particularly suitable for predicting extensive properties because its convolutional structure coincides with that of the covariance formula between sums of random variables. Using an active learning procedure, we demonstrate that the proposed method can achieve a mean absolute error of 0.62 ± 0.01 kcal/mol using as few as 2000 training samples on the QM7 dataset.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Data-driven deconvolution for large eddy simulations of Kraichnan turbulence

In this article, we demonstrate the use of artificial neural networks as optimal maps which are utilized for convolution and deconvolution of coarse-grained fields to account for sub-grid scale turbulence effects. Here, we demonstrate that an effective eddy-viscosity is predicted by our purely data-driven large eddy simulation framework without explicit utilization of phenomenological arguments. In addition, our data-driven framework precludes the knowledge of true sub-grid stress information during the training phase due to its focus on estimating an effective filter and its inverse so that grid-resolved variables may be related to direct numerical simulation data statistically. The proposed predictive framework is also combined with a statistical truncation mechanism for ensuring numerical realizability in an explicit formulation. Through this, we seek to unite structural and functional modeling strategies for modeling non-linear partial differential equations using reduced degrees of freedom. Both a priori and a posteriori results are shown for a two-dimensional decaying turbulence case in addition to a detailed description of validation and testing. A hyperparameter sensitivity study also shows that the proposed dual network framework simplifies learning complexity and is viable with exceedingly simple network architectures. Our findings indicate that the proposed framework approximates a robust and stable sub-grid closure which compares favorably to the Smagorinsky and Leith hypotheses for capturing the theoretical k -3 scaling in Kraichnan turbulence.

42 ENGINEERING↗

Ultrafast processing of pixel detector data with machine learning frameworks

Modern photon science performed at high repetition rate free-electron laser (FEL) facilities and beyond relies on 2D pixel detectors operating at increasing frequencies (towards 100 kHz at LCLS-II) and producing rapidly increasing amounts of data (towards TB/s). This data must be rapidly stored for offline analysis and summarized in real time for online feedback to the scientists. While at LCLS all raw data has been stored, at LCLS-II this would lead to a prohibitive cost; instead, enabling real time processing of pixel detector data (dark, gain, common mode, background, charge summing, subpixel position, photon counting, data summarization) allows reducing the size and cost of online processing, offline processing and storage by orders of magnitude while preserving full photon information. This could be achieved by taking advantage of the compressibility of sparse data typical for LCLS-II applications. Faced with a similar big data challenge a decade ago, computer vision stimulated revolutionary advances in machine learning hardware and software. We investigated if these developments are useful in the field of data processing for high speed pixel detectors and found that typical deep learning models and autoencoder architectures failed to yield useful noise reduction while preserving full photon information, presumably because of the very different statistics and feature sets in computer vision and radiation imaging. However, the raw performance of modern frameworks like Tensorflow inspired us to redesign in Tensorflow mathematically equivalent versions of the state-of-the-art, “classical” algorithms used at LCLS. The novel Tensorflow models resulted in elegant, compact and hardware agnostic code, gaining 1 to 2 orders of magnitude faster processing on an inexpensive consumer GPU, reducing by 3 orders of magnitude the projected cost of online analysis and compression without photon loss at LCLS-II. The novel Tensorflow models also enabled ongoing development of a pipelined hardware system expected to yield an additional 3 to 4 orders of magnitude speedup, necessary for meeting the data acquisition and storage requirements at LCLS-II, potentially enabling acquiring every single FEL pulse at full speed. Computer vision a decade ago was dominated by hand-crafted filters; their structure inspired the deep learning revolution resulting in modern deep convolutional networks; similarly, our novel Tensorflow filters provide inspiration for designing future deep learning models for ultrafast and efficient processing and classification of pixel detector images at FEL facilities.

47 OTHER INSTRUMENTATION↗

Dynamic absorption in optical pyrometry of hot spots in plastic-bonded triaminotrinitrobenzene [Optical Pyrometry of Shock-induced Hot Spots in Plastic-bonded Triaminotrinitrobenzene]

Herein we demonstrate a methodology for measuring optical pyrometry in environments which are disadvantageous for typical pyrometery applications by introducing additional fit parameters to account for absorption or emission which convolutes the thermal spectrum. Emission spectra from a plastic-bonded formulation of triaminotrinitrobenzene (TATB) shocked by 2-4 km s -1 impacts with Al flyer plates show significant deviations from graybody behavior. To extract reliable temperatures via optical pyrometry, we fit the spectra to a combination of a graybody and either a Gaussian absorption or emission spectrum. We found that the absorption needed to fit the data corresponds well with the known pressure-dependent absorption of TATB, and that the absorption model gives temperatures and emissivities in line with other explosives. By contrast, emission gives temperatures too low and emissivities that decrease as more material reacts. We conclude the nonthermal part of the spectrum is dominated by the absorption of unreacted TATB, and accurate pyrometry of TATB must use either our graybody plus absorption model or must limit the spectral range of observation to above 650 nm.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Mapping XANES spectra on structural descriptors of copper oxide clusters using supervised machine learning

Understanding the origins of enhanced reactivity of supported, subnanometer in size, metal oxide clusters is challenging due to the scarcity of methods capable to extract atomic-level information from the experimental data. Due to both the sensitivity of X-ray absorption near edge structure (XANES) spectroscopy to the local geometry around metal ions and reliability of theoretical spectroscopy codes for modeling XANES spectra, supervised machine learning approach has become a powerful tool for extracting structural information from the experimental spectra. Here, we present the application of this method to grazing incidence XANES spectra of size-selective Cu oxide clusters on flat support, measured in operando conditions of the methanation reaction. We demonstrate that the convolution neural network can be trained on theoretical spectra and utilized to “invert” experimental XANES data to obtain structural descriptors—the Cu–Cu coordination numbers. As a result, we were able to distinguish between different structural motifs (Cu 2 O-like and CuO-like) of Cu oxide clusters, transforming in reaction conditions, and reliably evaluate average cluster sizes, with important implications for the understanding of structure, composition, and function relationships in catalysis.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Machine learning surrogate models for Landau fluid closure

The first result of applying the machine/deep learning technique to the fluid closure problem is presented in this letter. As a start, three different types of neural networks (multilayer perceptron (MLP), convolutional neural network (CNN) and two-layer discrete Fourier transform (DFT) network) were constructed and trained to learn the well-known Hammett-Perkins Landau fluid closure in configuration space. We found that in order to train a well-preformed network, a minimum size of training data set is needed; MLP also requires a minimum number of neurons in the hidden layers equals to the degrees of freedom in Fourier space despite training data is fed in configuration space. Out of three models DFT performs the best for the clean data most likely due to the existence of nice Fourier expression for Hammett-Perkins closure but it is least robust with respect to input noise. Overall, with appropriate tuning and optimization, all three neural networks are able to accurately predict Hammett-Perkins closure and reproduce the inherit nonlocal feature, suggesting a promising path to calculate more sophisticated closures with the machine/deep learning technique.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Microparticle cloud imaging and tracking for data-driven plasma science

Oceans of image and particle track data encountered in plasma interactions with microparticle clouds motivate development and applications of machine-learning (ML) algorithms. A local-constant-velocity tracker, a Kohonen neural network or self-organizing map, the feature tracking kit, and U-Net are described and compared with each other for microparticle cloud datasets generated from exploding wires, dusty plasmas, and atmospheric plasmas. Particle density and the signal-to-noise ratio have been identified as two important factors that affect the tracking accuracy. Fast Fourier transform is used to reveal how U-Net, a deep convolutional neural network developed for non-plasma applications, achieves the improvements for noisy scenes. Viscous effects are revealed in the ballistic motions of the particles from the exploding wires and atmospheric plasmas. Subdiffusion of microparticles satisfying Δr 2 ∝t k (k=0.84±0.02) is obtained from the dusty plasma datasets. Microparticle cloud imaging and tracking, when enhanced with data and ML models, present new possibilities for plasma physics.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Development and modeling for a small-scale, rapidly heated high explosives initiation time (HEIT) experiment

Most small-scale assessments of explosive sensitivity, including the popular drop-weight impact test, convolute thermal and mechanical phenomena to the extent that it has been extremely challenging to decipher how an explosive ignites and propagates reactions. For instance, an impact generates heat through a range of dissipation mechanisms, which can in turn, depending on the reaction rates of the explosive, lead to chemical decomposition. To deconvolute the various contributions to the sub-shock initiation and propagation of explosive reactions, we describe the development and modeling of the High Explosives Initiation Time (HEIT) test - a new, small-scale, high-throughput experiment designed to rapidly heat milligram quantities of energetic materials confined within small diameter steel needles. Specifically, we have modeled and designed a 250 joule pulsed power system capable of rapidly delivering electrical current to the needles, resulting in rapid heat delivery to the sample. In conclusion, the energy deposition rate into the sample is controlled by different transmission line topologies. Modeling in COMSOL is performed to understand the energy required to heat up the explosive sample.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Equation of motion coupled-cluster cumulant approach for intrinsic losses in x-ray spectra

We present an equation of motion coupled cluster approach for calculating and understanding intrinsic inelastic losses in core level x-ray absorption spectra (XAS). The method is based on a factorization of the transition amplitude in the time-domain, which leads to a convolution of an effective one-body spectrum and the core-hole spectral function. The spectral function characterizes these losses in terms of shake-up excitations and satellites, and is calculated using a cumulant representation of the core-hole Green’s function that includes non-linear corrections. The one-body spectrum also includes orthogonality corrections that enhance the XAS at the edge.

Rehr, John J.↗