Search NASA⌕ Search

SEARCH · Search NASA

Results for “Experimental Methods”

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 1,333 records · Page 74

Loss Landscape Analysis for Reliable Quantized ML Models for Scientific Sensing

In this paper, we propose a method to perform empirical analysis of the loss landscape of machine learning (ML) models. The method is applied to two ML models for scientific sensing, which necessitates quantization to be deployed and are subject to noise and perturbations due to experimental conditions. Our method allows assessing the robustness of ML models to such effects as a function of quantization precision and under different regularization techniques -- two crucial concerns that remained underexplored so far. By investigating the interplay between performance, efficiency, and robustness by means of loss landscape analysis, we both established a strong correlation between gently-shaped landscapes and robustness to input and weight perturbations and observed other intriguing and non-obvious phenomena. Our method allows a systematic exploration of such trade-offs a priori, i.e., without training and testing multiple models, leading to more efficient development workflows. This work also highlights the importance of incorporating robustness into the Pareto optimization of ML models, enabling more reliable and adaptive scientific sensing systems.

Baldi, Tommaso [Pisa, Scuola Normale Superiore]↗

Perspectives on Prediction Variance and Bias in Developing, Assessing, and Comparing Experimental Designs

The vast majority of response surface methods used in practice to develop, assess, and compare experimental designs focus on variance properties of designs. Because response surface models only approximate the true unknown relationships, models are subject to bias errors as well as variance errors. Beginning with the seminal paper of Box and Draper (1959) and over the subsequent 50 years, methods that consider bias and mean-squared-error (variance and bias) properties of designs have been presented in the literature. However, these methods are not widely implemented in software and are not routinely used to develop, assess, and compare experimental designs in practice. Methods for developing, assessing, and comparing response surface designs that account for variance properties are reviewed. Brief synopses of publications that consider bias or mean-squared-error properties are provided. The difficulties and approaches for addressing bias properties of designs are summarized. Perspectives on experimental design methods that account for bias and/or variance properties and on future needs are presented.

97 MATHEMATICS AND COMPUTING↗

Electron density distribution in BaPb{sub 1-x}Sb{sub x}O{sub 3} superconducting oxides studied by double nuclear magnetic resonance methods

The effect of charge disorder on the formation of an inhomogeneous state of the electron system in the conduction band in BaPb{sub 1-x}Sb{sub x}O{sub 3} superconducting oxides is investigated experimentally by NMR methods. The NMR spectra of {sup 17}O are measured systematically, and the contributions from {sup 17}O atoms with different cation nearest surroundings are identified. It is found that microscopic regions with an elevated spin density of charge carriers are formed within two coordination spheres near antimony ions. Nuclei of the superconducting phase of the oxide (regions with an elevated antimony concentration) microscopically distributed over the sample are detected in compounds with x = 0.25 and 0.33. Experiments in which a double resonance signal of the spin echo of {sup 17}O-{sup 207}Pb and {sup 17}O-{sup 121}Sb are measured in the metal phase of BaPb{sub 1-x}Sb{sub x}O{sub 3} oxides are carried out for the first time. The constants of indirect heteronuclear spin-spin {sup 17}O-{sup 207}Pb interaction are determined as functions of the local Knight shift {sub 207}Ks. The estimates of the constants of the indirect interaction between the nuclei of the nearest neighbors (O-Pb and Pb-Pb atoms) and analysis of evolution of the NMR spectra of {sup 17}O upon a change in the antimony concentration are convincing evidence in favor of the development of a microscopically inhomogeneous state of the electron system in the metal phase of BaPb{sub 1-x}Sb{sub x}O{sub 3} oxides.

36 MATERIALS SCIENCE↗

A method to determine the onset voltage of single and arrays of electrospray emitters

This paper reports on an accurate and rapid method to compute the onset voltage of a single or an array of electrospray emitters with complex geometries and on the correlation of the simulation with experimental data. This method permits the exact determination of the onset voltage based only on the surface tension of the sprayed liquid and on the emitter geometry. The approach starts by determining the voltage at which electrostatic forces and surface tension forces are equal for a sharpening conic surface at the tip of a capillary as a function of the apex radius of the liquid. By tracing the curve of this computed equilibrium voltage as a function of the apex radius, the onset voltage for a liquid surface with the Taylor half-angle of 49.3 deg. or larger can be determined. For smaller cone half-angles the method is only applicable to ionic sprays as an approximate knowledge of the critical field for ion emission is necessary. The combination of analytical models and finite element tools used to compute the necessary parameters is described. The method is validated on a complex microelectromechanical system emitter geometry as well as on a linear array of electrospray emitters. Finally an empirical model of the behavior of the electric field near the apex of a conic surface with asymptotes at a fixed half-angle is proposed, which allows establishing a simple method for onset voltage determination.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Imaging particle collision data for event classification using machine learning

We propose a method to organize experimental data from particle collision experiments in a general format which can enable a simple visualisation and effective classification of collision data using machine learning techniques. The method is based on sparse fixed-size matrices with single- and two-particle variables containing information on identified particles and jets. As a result, we illustrate this method using an example of searches for new physics at the LHC experiments.

47 OTHER INSTRUMENTATION↗

DeepHyper: Asynchronous Hyperparameter Search for Deep Neural Networks

Hyperparameters employed by deep learning (DL) methods play a substantial role in the performance and reliability of these methods in practice. Unfortunately, finding performance-optimizing hyperparameter settings is a notoriously difficult task. Hyperparameter search methods typically have limited production-strength implementations or do not target scalability within a highly parallel machine, portability across different machines, experimental comparison between different methods, and tighter integration with workflow systems. In this paper, we present DeepHyper, a Python package that provides a common interface for the implementation and study of scalable hyperparameter search methods. It adopts the Balsam workflow system to hide the complexities of running large numbers of hyperparameter configurations in parallel on high-performance computing (HPC) systems. We implement and study asynchronous model-based search methods that consist of sampling a small number of input hyperparameter configurations and progressively fitting surrogate models over the input-output space until exhausting a user-defined budget of evaluations. We evaluate the efficacy of these methods relative to approaches such as random search, genetic algorithms, Bayesian optimization, and hyperband on DL benchmarks on CPU- and GPU-based HPC systems.

Balaprakash, Prasanna↗

Capacitor Voltage Ripple Reduction of A 4-Level Hybrid Clamped Converter Using Switching State Redundancy for Medium-Voltage Drive

Abstract—A four-level Hybrid Clamped Converter (4LHCC) is a novel promising DC/AC multilevel converter for high-efficiency medium-voltage drives. This paper presents a new control method of the 4L-HCC using the switching state redundancy to reduce the DC-link capacitor voltage ripples at low frequencies (<=30 Hz). The new switching state selection algorithm and flying capacitor balancing method are introduced and analyzed.Both of a3.6 kV 1MW 4L-HCC simulation model and a SiC-based 4L-HCC prototype is established to verify the proposed method. Simulation and experimental result shows that the maximum DC-link capacitor ripple can be reduced effectively from over 50% to less than 5% at low frequencies with the new method. Meanwhile, load voltage and currentdistortion caused by the excessive capacitor rippleisalso mitigatedsignificantly.

hybrid clamped converter, motor drive, voltage bal↗

An investigation of geometrically necessary dislocations and back stress in large grained tantalum via EBSD and CPFEM

This study explores the evolution of GNDs and their effects on back stress through experimental and computational methods. Four large-grained tantalum tensile specimens were strained in uniaxial tension, electron backscatter diffraction (EBSD) data were collected, and geometrically necessary dislocation (GND) maps of the four specimens in the unloaded state were produced. EBSD-based GND maps revealed several types of features with high GND content which caused back stress in the specimens. Correlations between five geometrically-based grain boundary (GB) transmission factors and the GB GND content were evaluated, and statistically significant correlations were found for transmission factors based on Livingston and Chalmer’s N factor, Werner and Prantl’s slip transfer number, and GB misorientation. The sign of individual components of the Nye tensor were used to visually and quantitatively identify clustering of GNDs of the same sign, thus giving additional evidence of increasing back stress due to deformation. Deformation of one of the specimens was simulated using multiple CPFEM based modeling approaches and predicted stress-strain responses are compared. The super dislocation model (SD model) — a crystal plasticity finite element method (CPFEM) which incorporates elastic dislocation interactions — was able to isolate impact of back stress on the overall flow stress. The SD model predicted correct stresses when compared with experimental data; however, when the elastic interactions in the SD model were turned off, stress predictions were 25% too low. Thus, demonstrating the importance of incorporating back stress into the model.

42 ENGINEERING↗

Experimental energy-dependent nuclear spin distributions

A new method is proposed to determine the energy-dependent spin distribution in experimental nuclear-level schemes. This method compares various experimental and calculated moments in the energy-spin plane to obtain the spin-cutoff parameter {sigma} as a function of mass A and excitation energy using a total of 7202 levels with spin assignment in 227 nuclei between F and Cf. A simple formula, {sigma}{sup 2}=0.391 A{sup 0.675}(E-0.5Pa{sup '}){sup 0.312}, is proposed up to about 10 MeV that is in very good agreement with experimental {sigma} values and is applied to improve the systematics of level-density parameters.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Predictive Direct DC-Link Control for 7.2 kV Three-Port Low-Inertia Solid-State Transformer With Active Power Decoupling

Promising for applications including renewable energy and electric vehicle fast charging, a medium-voltage (MV) solid-state transformer (SST) typically has multiple modules series stacked, which requires the modules to be single phase. One critical issue of the single-phase SST is the double-line-frequency power ripple. Traditionally, large passives are used to buffer the ripple, resulting in significantly increased volume and cost. This article for the first time proposes active power decoupling (APD) for the SST and a predictive direct DC-link control method, using a current-source single-stage SST as an example. An electrolytic capacitor-less buffer port is used to absorb the ripple, which tolerates up to a 30% voltage ripple for small capacitance. The APD SST control is challenging. First, the current-source SST realizes isolation, DC link, and multiport in a single stage. Second, different from conventional SSTs or APD converters, the low-inertia DC link in this SST has a 40% switching ripple, which is difficult to stabilize. To address the control challenge, a direct DC-link control architecture and a predictive control method are proposed. Simulation and experimental results verify the proposed control method on an MV SiC modular soft-switching solid-state transformer (M-S4T) prototype. Here, the proposed APD reduces the SST volume by 53.8%. Importantly, the proposed concept is not limited to the M-S4T but is generic to other SSTs or APD converters.

14 SOLAR ENERGY↗

Learning statistical correlation for fast prostate registration in image-guided radiotherapy

Purpose: In adaptive radiation therapy of prostate cancer, fast and accurate registration between the planning image and treatment images of the patient is of essential importance. With the authors' recently developed deformable surface model, prostate boundaries in each treatment image can be rapidly segmented and their correspondences (or relative deformations) to the prostate boundaries in the planning image are also established automatically. However, the dense correspondences on the nonboundary regions, which are important especially for transforming the treatment plan designed in the planning image space to each treatment image space, are remained unresolved. This paper presents a novel approach to learn the statistical correlation between deformations of prostate boundary and nonboundary regions, for rapidly estimating deformations of the nonboundary regions when given the deformations of the prostate boundary at a new treatment image. Methods: The main contributions of the proposed method lie in the following aspects. First, the statistical deformation correlation will be learned from both current patient and other training patients, and further updated adaptively during the radiotherapy. Specifically, in the initial treatment stage when the number of treatment images collected from the current patient is small, the statistical deformation correlation is mainly learned from other training patients. As more treatment images are collected from the current patient, the patient-specific information will play a more important role in learning patient-specific statistical deformation correlation to effectively reflect prostate deformation of the current patient during the treatment. Eventually, only the patient-specific statistical deformation correlation is used to estimate dense correspondences when a sufficient number of treatment images have been acquired from the current patient. Second, the statistical deformation correlation will be learned by using a multiple linear regression (MLR) model, i.e., ridge regression (RR) model, which has the best prediction accuracy than other MLR models such as canonical correlation analysis (CCA) and principal component regression (PCR). Results: To demonstrate the performance of the proposed method, we first evaluate its registration accuracy by comparing the deformation field predicted by our method with the deformation field estimated by the thin plate spline (TPS) based correspondence interpolation method on 306 serial prostate CT images of 24 patients. The average predictive error on the voxels around 5 mm of prostate boundary is 0.38 mm for our method of RR-based correlation model. Also, the corresponding maximum error is 2.89 mm. We then compare the speed for deformation interpolation by different methods. When considering the larger region of interest (ROI) with the size of 512 x 512 x 61, our method takes 24.41 seconds to interpolate the dense deformation field while TPS method needs 6.7 minutes; when considering a small ROI (surrounding prostate) with size of 112 x 110 x 93, our method takes 1.80 seconds, while TPS method needs 25 seconds. Conclusions: Experimental results show that the proposed method can achieve much faster registration speed yet with comparable registration accuracy, compared to the TPS-based correspondence (or deformation) interpolation approach.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Masking of a High-Reflectivity Sphere by a Layer with a Random Quasi-Zero Refractive Index

The method of masking of a large-radius sphere made of an arbitrary material by a masking coating with a random quasi-zero refractive index is substantiated theoretically and experimentally. In this method of masking, the sphere is first coated with a mirror layer (e.g., high-reflectivity silver) and then with a masking layer with a thickness much smaller than the radius of the sphere. The fraction of the external radiation intensity bending around the sphere at the points of observation near the surface of the sphere and the extinction cross section of the sphere with the masking coating for points of observation far away from the sphere are calculated. It is shown that extinction cross section Q of the sphere with a masking coating is much smaller than 2D, where D is the geometrical cross section of the sphere; this corresponds to the masking effect for the sphere in the wavelength range of at least 450–1200 nm in the transparency region of the masking coating.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Computer-aided Abnormality Detection in Chest Radiographs in a Clinical Setting via Domain-adaptation

Deep learning (DL) models are being deployed at medical centers to aid radiologists for diagnosis of lung conditions from chest radiographs. Such models are often trained on a large volume of publicly available labeled radiographs. These pre-trained DL models’ ability to generalize in clinical settings is poor because of the changes in data distributions between publicly available and privately held radiographs. In chest radiographs, the heterogeneity in distributions arises from the diverse conditions in X-ray equipment and their configurations used for generating the images. In the machine learning community, the challenges posed by the heterogeneity in the data generation source is known as domain shift, which is a mode shift in the generative model. In this work, we introduce a domain-shift detection and removal method to overcome this problem. Our experimental results show the proposed method’s effectiveness in deploying a pre-trained DL model for abnormality detection in chest radiographs in a clinical setting.

Dubey, Abhishek↗

Comparative Omics-Driven Genome Annotation Refinement: Application across Yersiniae

Genome sequencing continues to be a rapidly evolving technology, yet most downstream aspects of genome annotation pipelines remain relatively stable or are even being abandoned. To date, the perceived value of manual curation for genome annotations is not offset by the real cost and time associated with the process. In order to balance the large number of sequences generated, the annotation process is now performed almost exclusively in an automated fashion for most genome sequencing projects. One possible way to reduce errors inherent to automated computational annotations is to apply data from 'omics' measurements (i.e. transcriptional and proteomic) to the un-annotated genome with a proteogenomic-based approach. This approach does require additional experimental and bioinformatics methods to include omics technologies; however, the approach is readily automatable and can benefit from rapid developments occurring in those research domains as well. The annotation process can be improved by experimental validation of transcription and translation and aid in the discovery of annotation errors. Here the concept of annotation refinement has been extended to include a comparative assessment of genomes across closely related species, as is becoming common in sequencing efforts. Transcriptomic and proteomic data derived from three highly similar pathogenic Yersiniae (Y. pestis CO92, Y. pestis pestoides F, and Y. pseudotuberculosis PB1/+) was used to demonstrate a comprehensive comparative omic-based annotation methodology. Peptide and oligo measurements experimentally validated the expression of nearly 40% of each strain's predicted proteome and revealed the identification of 28 novel and 68 previously incorrect protein-coding sequences (e.g., observed frameshifts, extended start sites, and translated pseudogenes) within the three current Yersinia genome annotations. Gene loss is presumed to play a major role in Y. pestis acquiring its niche as a virulent pathogen, thus the discovery of many translated pseudogenes underscores a need for functional analyses to investigate hypotheses related to divergence. Refinements included the discovery of a seemingly essential ribosomal protein, several virulence-associated factors, and a transcriptional regulator, among other proteins, most of which are annotated as hypothetical, that were missed during annotation.

59 BASIC BIOLOGICAL SCIENCES↗

Experimental Investigation of Subcritical Neutron and Gamma Noise Methods at the Seven Percent Critical Experiment (7uPCX)

As part of a collaborative international effort organized by the Lawrence Livermore National Laboratory (LLNL), with key participants from the Institut de radioprotection et de sûreté nucléaire (IRSN), Los Alamos National Laboratory (LANL), and the Sandia National Laboratories (SNL), a series of high-multiplication subcritical neutron and gamma noise measurements was planned and executed. The primary aim of this article was to advance detector technology, assess the validity of gamma noise for subcriticality measurements, and nuclear criticality safety, focusing on collecting list-mode or time-series data from various reactor configurations with multiplication values ranging from 20 to 310. This comprehensive dataset enabled a detailed comparative analysis of multiple detector systems and the results of both neutron and gamma noise measurements. In this work, we focus on experimentally comparing the results from neutron and gamma noise measurements. We note good agreement between estimations of the prompt neutron decay constant and demonstrate the effects of changing reactor geometry on the efficiency of the differing methods. Our results agree well with independent experimental measurements and simulations performed by LLNL.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Electronic structures and optical properties of wurtzite type LiBSe{sub 2} (B=Al, Ga, In): A first-principles study

The electronic structures of three wurtzite type isostructural compounds LiBSe{sub 2} (B=Al, Ga, In) are studied by the density functional theory (DFT). The results reveal that the presence of Li cations has direct influence on neither the band gaps (Eg) nor the bonding levels, but plays an important role in the stabilization of the structures. The band structures and densities of states (DOS) are analyzed in detail, and the band gaps of LiBSe{sub 2} adhere to the following trend Eg{sub (LiAlSe2)}>Eg{sub (LiGaSe2)}>Eg{sub (LiInSe2)}, which is in agreement with the decrease of the bond energy of the corresponding Se 4p-B s antibonding orbitals. The role of the active s electrons of B element on the band gaps is also discussed. Finally, the optical properties are predicted, and the results would be a guide to understand the experiments. - Graphical abstract: The electronic structures and optical properties of wurtzite type LiBSe{sub 2} (B=Al, Ga, In) have been studied by the DFT calculations. And the correlation of the electronegative of B element and the band gap decrease-trend are discussed. The comparison between different calculation methods and the experimental results is presented.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Subtractive renormalization of the NN interaction in chiral effective theory up to next-to-next-to-leading order: S waves

We extend our subtractive-renormalization method to evaluate the {sup 1}S{sub 0} and {sup 3}S{sub 1}-{sup 3}D{sub 1} NN-scattering phase shifts up to next-to-next-to-leading order (NNLO) in chiral effective theory. We show that, if energy-dependent contact terms are employed in the NN potential, the {sup 1}S{sub 0} phase shift can be obtained by carrying out two subtractions on the Lippmann-Schwinger equation. These subtractions use knowledge of the the scattering length and the {sup 1}S{sub 0} phase shift at a specific energy to eliminate the low-energy constants in the contact interaction from the scattering equation. For the J=1 coupled channel, a similar renormalization can be achieved by three subtractions that employ knowledge of the {sup 3}S{sub 1}scattering length, the {sup 3}S{sub 1} phase shift at a specific energy, and the {sup 3}S{sub 1}-{sup 3}D{sub 1} generalized scattering length. In both channels a similar method can be applied to a potential with momentum-dependent contact terms, except that in that case one of the subtractions must be replaced by a fit to one piece of experimental data. This method allows the use of arbitrarily high cutoffs in the Lippmann-Schwinger equation. We examine the NNLO S-wave phase shifts for cutoffs as large as 19 GeV and show that the presence of linear energy dependence in the NN potential creates spurious poles in the scattering amplitude. In consequence the results are in conflict with empirical data over appreciable portions of the considered cutoff range. We also identify problems with the use of cutoffs greater than 1 GeV when momentum-dependent contact interactions are employed. These problems are ameliorated, but not eliminated, by the use of spectral-function regularization for the two-pion exchange part of the NN potential.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Experimental Evaluation of Deformation and Fracture Mechanisms in Highly Irradiated Austenitic Steels

The present report documents recent experimental results of analysis using scanning electron microscopy/electron backscatter diffraction (SEM-EBSD) of plastic deformation mechanisms and strain localization phenomena in austenitic steels irradiated by neutrons. Experiments were performed with specimens irradiated to 125 dpa and, additionally, with specimens that experienced radiation-induced swelling up to 3%. Section 1 briefly analyzes the deformation localization in irradiated steels and its consequences on the material performance. The section describes the advantages and importance of the SEM-EBSD approach combined with in situ mechanical testing capability. Section 2 briefly introduces the experimental tools and methods (i.e., SEM/EBSD in situ tensile frame, electric discharge machine to manufacture irradiated specimens) and describes the investigated materials (i.e., element composition, irradiation conditions, and general microstructure). Section 3 describes the key experimental results and provides a brief analysis and comparison with the datasets obtained earlier within the same task (i.e., low-dose specimens). The discussion focuses on EBSD microstructure maps with strain localization features, misorientation evolution as a function of strain, and observed deformation mechanisms. Section 4 evaluates data collected in recent years on highly irradiated steel and estimates the possible misorientation evolution under irradiation. The section introduces and discusses the concept of in-service-induced damage as an irradiation-assisted stress-corrosion cracking precursor. Section 5 summarizes the work performed. As expected, the present work results are beneficial for exploring and understanding degradation mechanisms in highly irradiated in-core materials found in light water reactors after long-term in-service life.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗