Search NASA⌕ Search

SEARCH · Search NASA

Results for “tensor 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 1,171 records · Page 65

WE-A-17A-11: Implanted Brachytherapy Seed Movement Due to Transrectal Ultrasound Probe-Induced Prostate Deformation

Purpose: To characterize the movement of implanted brachytherapy seeds due to transrectal ultrasound probe-induced prostate deformation and to estimate the effects on prostate dosimetry. Methods: Implanted probe-in and probe-removed seed distributions were reconstructed for 10 patients using C-arm fluoroscopy imaging. The prostate was delineated on ultrasound and registered to the fluoroscopy seeds using a visible subset of seeds and residual needle tracks. A linear tensor and shearing model correlated the seed movement with position. The seed movement model was used to infer the underlying prostate deformation and to simulate the prostate contour without probe compression. Changes in prostate and surrogate urethra dosimetry were calculated. Results: Seed movement patterns reflecting elastic decompression, lateral shearing, and rectal bending were observed. Elastic decompression was characterized by anterior-posterior expansion and superior-inferior and lateral contractions. For lateral shearing, anterior movement up to 6 mm was observed for extraprostatic seeds in the lateral peripheral region. The average intra-prostatic seed movement was 1.3 mm, and the residual after linear modeling was 0.6 mm. Prostate D90 increased by 4 Gy on average (8 Gy max) and was correlated with elastic decompression. For selected patients, lateral shearing resulted in differential change in D90 of 7 Gy between anterior and posterior quadrants, and increase in whole prostate D90 of 4 Gy. Urethra D10 increased by 4 Gy. Conclusion: Seed movement upon probe removal was characterized. The proposed model captured the linear correlation between seed movement and position. Whole prostate dose coverage increased slightly, due to the small but systematic seed movement associated with elastic decompression. Lateral shearing movement increased dose coverage in the anterior-lateral region, at the expense of the posterior-lateral region. The effect on whole prostate D90 was smaller due to the subset of peripheral seeds involved, but lateral shearing movement can have greater consequences for local dose coverage.

60 APPLIED LIFE SCIENCES↗

Geometry-aware training of factorized layers in tensor Tucker format

Reducing parameter redundancies in neural network architectures is crucial for achieving feasible computational and memory requirements during train and inference of large networks. Given its easy implementation and flexibility, one promising approach is layer factorization, which reshapes weight tensors into a matrix format and parameterizes it as the product of two rank-r matrices. However, this family of approaches often requires an initial full-model warm-up phase, prior knowledge of a feasible rank, and it is sensitive to parameter initialization.In this work, we introduce a novel approach to train the factors of a Tucker decomposition of the weight tensors. Our training proposal proves to be optimal in locally approximating the original unfactorized dynamics and stable for the initialization. Furthermore, the rank of each mode is dynamically updated during training.We provide a theoretical analysis of the algorithm, showing convergence, approximation and local descent guarantees. The method's performance is further illustrated through a variety of experiments, showing remarkable training compression rates and comparable or even better performance than the full baseline and alternative layer factorization strategies.

Zangrando, Emanuele [Gran Sasso Science Institute ↗

Sparsity of Radiating Characteristic Modes on Infinite Periodic Structures

Characteristic modes on infinite periodic structures are studied using spectral dyadic Green’s functions. This formulation demonstrates that, in contrast to the modal analysis of finite structures, the number of radiating characteristic modes is limited by unit cell size and incident wave vector (i.e., scan angle or phase shift per unit cell). Here, the reflection tensor is decomposed into modal contributions from radiating modes, indicating that characteristic modes are a predictably sparse basis in which to study reflection phenomena.

42 ENGINEERING↗

A continuum dislocation-based model of wedge microindentation of single crystals

Recent Electron Backscatter Diffraction (EBSD) experiments have revealed the emergence of heterogeneous dislocation microstructures forming under a wedge indenter in fcc crystals, where micro-meter dislocation patterns challenge the predictions of traditional models of plasticity. In order to explain the formation of these features and develop a relationship between the force-displacement curve and the dislocation substructure, here we present a model of wedge indentation based on the continuum theory of dislocations. The model accounts for large deformation kinematics through the multiplicative split of the deformation gradient tensor, where the incompatible plastic component of deformation results from the flux of dislocations on different and interacting slips systems. Constitutive equations for dislocation fluxes are determined from a dissipative variational principle. As a result, each dislocation density satisfies an initial-boundary value problem with convective-diffusive character, which is coupled to the macroscopic stress and displacement fields governing the deformation process. Solution to the self-consistent continuum formulation is found using the finite element method. Computer simulations mimic the experimental conditions of wedge micro-indentation experiments into Ni single-crystals used by Kysar et al. (2010a). A comparison of overall dislocation density distribution and macroscopic mechanical response shows good overall agreement with the experimental results in terms of the detailed features of dislocation patterns and lattice rotations as well as the macroscopic force-displacement response.

36 MATERIALS SCIENCE↗

Bringing Trimmed Serendipity Methods to Computational Practice in Firedrake

We present an implementation of the trimmed serendipity finite element family, using the open-source finite element package Firedrake. The new elements can be used seamlessly within the software suite for problems requiring H 1 , H (curl), or H (div)-conforming elements on meshes of squares or cubes. To test how well trimmed serendipity elements perform in comparison to traditional tensor product elements, we perform a sequence of numerical experiments including the primal Poisson, mixed Poisson, and Maxwell cavity eigenvalue problems. Overall, we find that the trimmed serendipity elements converge, as expected, at the same rate as the respective tensor product elements, while being able to offer significant savings in the time or memory required to solve certain problems.

97 MATHEMATICS AND COMPUTING↗

Mixed-Precision S/DGEMM Using the TF32 and TF64 Frameworks on Low-Precision AI Tensor Cores

Using NVIDIA graphics processing units (GPUs) equipped with Tensor Cores has enabled the significant acceleration of general matrix multiplication (GEMM) for applications in machine learning (ML) and artificial intelligence (AI) and in high-performance computing (HPC) generally. The use of such power-efficient, specialized accelerators can provide a performance increase between 8 × and 20 ×, albeit with a loss in precision. However, a high level of precision is required in many large scientific and HPC applications, and computing in single or double precision is still necessary for many of these applications to maintain accuracy. Fortunately, mixed-precision methods can be employed to maintain a higher level of numerical precision while also taking advantage of the performance increases from computing with lower-precision AI cores. With this in mind, we extend the state of the art by using NVIDIA’s new TF32 framework. This new framework not only burdens some constraints of the previous frameworks, such as costly 32 16-bit castings but also provides an equivalent precision and performance by using a much simpler approach. We also propose a new framework called TF64 that attempts double-precision arithmetic with low-precision Tensor Cores. Although this framework does not exist yet, we validated the correctness of this idea and achieved an equivalent of 64-bit precision on 32-bit hardware.

Valero Lara, Pedro↗

Kinetic Effects on Resistive Wall Mode Stability

Two approaches for studying the damping of resistive wall modes due to wave particle resonant interactions are discussed. One approach uses the eigenfunction from an ideal MHD code combined with the resonant particle damping calculated from a drift-kinetic {delta}f- method formulation. This perturbative approach treats the wave-particle interaction precisely, but does not include the back-effect of the kinetic terms on the eigenfunction structure. In the alternate non-perturbative approach the kinetic terms are included in an MHD description via the pressure tensor. This non-perturbative approach includes the resonances due to the bounce and precessional drifts subject to certain approximations. Comparisons between the two approaches and conclusions on the dominant stabilizing mechanisms are presented.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Angular correlations in the inelastic-scattering process {sup 24}Mg(d, d{gamma}){sup 24}Mg at E{sub d} = 15.3 MeV

The double-differential cross sections for the reaction {sup 24}Mg(d, d{gamma}{sub 1.369}){sup 24}Mg at the projectiledeuteron energy of E{sub d} = 15.3 MeV weremeasured for deuteron emission angles in the forward hemisphere. All even spin-tensor components of the density matrix for the 2{sup +} state of the {sup 24}Mg nucleus at 1.369 MeV were reconstructed, and its orientation properties were determined. These experimental results were compared with the results of calculations based on various versions of the coupled-channel method.

74 ATOMIC AND MOLECULAR PHYSICS↗

CMB-S4: Foreground-cleaning Pipeline Comparison for Measuring Primordial Gravitational Waves

We compare multiple foreground-cleaning pipelines for estimating the tensor-to-scalar ratio, r, using simulated maps of the planned CMB-S4 experiment within the context of the South Pole Deep Patch. To evaluate robustness, we analyze bias and uncertainty on r across various foreground suites using map-based simulations. The foreground-cleaning methods include: a parametric maximum likelihood approach applied to auto- and cross-power spectra between frequency maps; a map-based parametric maximum-likelihood method; and a harmonic-space internal linear combination using frequency maps. We summarize the conceptual basis of each method to highlight their similarities and differences. To better probe the impact of foreground residuals, we implement an iterative internal delensing step, leveraging a map-based pipeline to generate a lensing B-mode template from the large aperture telescope frequency maps. Our results show that the performance of the three approaches is comparable for simple and intermediate-complexity foregrounds, with σ(r) ranging from 3–5 ×10 −4 . However, biases at the 1σ–2σ level appear when analyzing more complex forms of foreground emission. By extending the baseline pipelines to marginalize over foreground residuals, we demonstrate that contamination can be reduced to within statistical uncertainties, albeit with a pipeline-dependent impact on σ(r), which translates to a detection significance between 2σ and 4σ for an input value of r = 0.003. These findings suggest varying levels of maturity among the tested pipelines, with the auto- and cross-spectra-based approach demonstrating the best stability and overall performance. Moreover, given the extremely low noise levels, mutual validation of independent foreground-cleaning pipelines is essential to ensure the robustness of any potential detection.

astronomy data analysis↗

Temperature dependent behavior of localized and delocalized electrons in nitrogen-doped 6H SiC crystals as studied by electron spin resonance

We have studied the temperature behavior of the electron spin resonance (ESR) spectra of nitrogen (N) donors in n-type 6H SiC crystals grown by Lely and sublimation sandwich methods (SSM) with donor concentration of 10{sup 17 }cm{sup −3} at T = 60–150 K. A broad signal in the ESR spectrum was observed at T ≥ 80 K with Lorentzian lineshape and g{sub ||} = 2.0043(3), g{sub ⊥} = 2.0030(3), which was previously assigned in the literature to the N donors in the 1s(E) excited state. Based on the analysis of the ESR lineshape, linewidth and g-tensor we attribute this signal to the conduction electrons (CE). The emergence of the CE ESR signal at T > 80 K was explained by the ionization of electrons from the 1s(A{sub 1}) ground and 1s(E) excited states of N donors to the conduction band while the observed reduction of the hyperfine (hf) splitting for the N{sub k1,k2} donors with the temperature increase is attributed to the motional narrowing effect of the hf splitting. The temperature dependence of CE ESR linewidth is described by an exponential law (Orbach process) with the activation energy corresponding to the energy separation between 1s(A{sub 1}) and 1s(E) energy levels for N residing at quasi-cubic sites (N{sub k1,k2}). The theoretical analysis of the temperature dependence of microwave conductivity measured by the contact-free method shows that due to the different position of the Fermi level in two samples the ionization of free electrons occurs from the energy levels of N{sub k1,k2} donors in Lely grown samples and from the energy level of N{sub h} residing at hexagonal position in 6H SiC grown by SSM.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Non-Gaussian structure of B-mode polarization after delensing

The B-mode polarization of the cosmic microwave background on large scales has been considered as a probe of gravitational waves from the cosmic inflation. Ongoing and future experiments will, however, suffer from contamination due to the B-modes of non-primordial origins, one of which is the lensing induced B-mode polarization. Subtraction of the lensing B-modes, usually referred to as delensing, will be required for further improvement of detection sensitivity of the gravitational waves. In such experiments, knowledge of statistical properties of the B-modes after delensing is indispensable to likelihood analysis particularly because the lensing B-modes are known to be non-Gaussian. In this paper, we study non-Gaussian structure of the delensed B-modes on large scales, comparing it with that of the lensing B-modes. In particular, we investigate the power spectrum correlation matrix and the probability distribution function (PDF) of the power spectrum amplitude. Assuming an experiment in which the quadratic delensing is an almost optimal method, we find that delensing reduces correlations of the lensing B-mode power spectra between different multipoles, and that the PDF of the power spectrum amplitude is well described as a normal distribution function with a variance larger than that in the case of a Gaussian field. These features are well captured by an analytic model based on the 4th order Edgeworth expansion. As a consequence of the non-Gaussianity, the constraint on the tensor-to-scalar ratio after delensing is degraded within approximately a few percent, which depends on the multipole range included in the analysis.

79 ASTRONOMY AND ASTROPHYSICS↗

Casimir effect for parallel plates in de Sitter spacetime

The Wightman function and the vacuum expectation values of the field squared and of the energy-momentum tensor are obtained, for a massive scalar field with an arbitrary curvature coupling parameter, in the region between two infinite parallel plates, on the background of de Sitter spacetime. The field is prepared in the Bunch-Davies vacuum state and is constrained to satisfy Robin boundary conditions on the plates. For the calculation, a mode-summation method is used, supplemented with a variant of the generalized Abel-Plana formula. This allows one to explicitly extract the contributions to the expectation values, which come from each single boundary, and to expand the second-plate-induced part in terms of exponentially convergent integrals. Several limiting cases of interest are then studied. Moreover, the Casimir forces acting on the plates are evaluated, and it is shown that the curvature of the background spacetime decisively influences the behavior of these forces at separations larger than the curvature scale of de Sitter spacetime. In terms of the curvature coupling parameter and the mass of the field, two very different regimes are realized, which exhibit monotonic and oscillatory behavior of the vacuum expectation values, respectively. The decay of the Casimir force at large plate separation is shown to be power law (monotonic or oscillating), with independence of the value of the field mass.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Electromagnetic field of a charge traveling into an anisotropic medium

We analyze the electromagnetic field generated by a point charge intersecting the interface between vacuum and a nonmagnetic anisotropic medium with a plasma-type dispersion of the dielectric permittivity tensor. After penetrating the medium, the charge moves along its main axis. The total field is presented as a sum of a self-field (i.e., a charge field in a corresponding unbounded medium) and a scattered field associated with the boundary influence. We show that the self-field in the considered anisotropic medium is divided into a quasistatic field and a wave field (the so-called 'plasma trace' is absent in the case under consideration). Under certain conditions, the Vavilov-Cherenkov radiation generated in the medium is reversed (i.e., the energy flux density vector forms an obtuse angle with the direction of the charge motion). Accordingly, so-called reversed Cherenkov-transition radiation (RCTR) can be generated. We analytically and numerically investigate both the scattered field and the total one, and we show that RCTR exists in the vacuum region if the charge velocity exceeds a certain threshold value associated with total internal reflection. Computations of the Fourier harmonics of the field as well as the total field itself demonstrate that RCTR in vacuum can be a dominant effect. Some properties of RCTR can be useful for diagnostics of particle bunches and determination of medium characteristics.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Effects from Spin–Orbit Coupling on Electron–Nucleus Hyperfine Coupling Calculated at the Restricted Active Space Level for Kramers Doublets

Calculations of electron–nucleus hyperfine coupling were implemented at the restricted active space state interaction (RASSI) level to treat spin–orbit (SO) coupling, based on scalar relativistic restricted active space wave functions. The current implementation is suitable for light atomic systems, for light ligand atoms in heavy metal complexes, and for spin–orbit coupling-induced hyperfine coupling of heavy atoms if the unpaired electrons are described by orbitals with high angular momentum. Spin polarization is reasonably well treated by allowing one hole and one electron in a window of active orbitals (‘ras1’, ‘ras3’) surrounding the principal active space (‘ras2’). A benchmark set of Kramers doublet states of molecules with light and heavy atoms is used to evaluate the approach and verify the implementation. For NpF 6 , the impact of SO coupling on the Np and F hyperfine coupling tensors is investigated in detail. It is demonstrated that the Np hyperfine coupling is strongly dominated by SO effects, that there is a large SO effect on the F hyperfine tensor components, and that SO coupling causes the fluorine dipolar term to acquire an isotropic component.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Calculating Trajectories Associated With Solute Transport in a Heterogeneous Medium

Here, we present a trajectory-based technique for calculating solute transport in a porous medium that has several advantages over existing methods. Unlike streamlines, the extended trajectories are influenced by each of the important parameters governing transport, including molecular diffusion and transverse dispersion. The approach is complete and does not require any additional techniques, such as operator splitting or particle tracking, in order to account for the full dispersion tensor. The semianalytic expressions make it clear how the flow field, the concentration distribution, and the dispersion tensor contribute to the velocity field of an injected solute. The equations are valid for an arbitrary porous medium, including those with rapid spatial variations in properties, overcoming limitations faced by previous approaches based upon asymptotic techniques. A test on a layered model with sharp boundaries indicates that the extended trajectories are compatible with the results of a numerical simulator and differ from streamlines. We also describe a new form of the dispersion tensor that incorporates a known asymmetry. The trajectories indicate that the modifications of the dispersion tensor lead to more focused transport within regions of high conductivity. The trajectories are used to define a semianalytic relationship between solute travel times and variations in solute velocities along a path that may be used for tomographic imaging. In an application to the injection of a radioactive tracer into a Berea sandstone core, monitored using micropositron emission tomographic (micro-PET) observations, the sensitivities are used to map the spatial variations of permeability within the core.

42 ENGINEERING↗

A self-consistent hybrid model of kinetic striations in low-current argon discharges

A self-consistent hybrid model of standing and moving striations was developed for low-current DC discharges in noble gases. We introduced the concept of surface diffusion in phase space ( r , u ) (where u denotes the electron kinetic energy) described by a tensor diffusion in the nonlocal Fokker–Planck kinetic equation for electrons in the collisional plasma. Electrons diffuse along surfaces of constant total energy ε = u - eφ ( r ) between energy jumps in inelastic collisions with atoms. Numerical solutions of the 1d1u kinetic equation for electrons were obtained by two methods and coupled to ion transport and Poisson solver. We studied the dynamics of striation formation in Townsend and glow discharges in argon gas at low discharge currents using a two-level excitation-ionization model and a ‘full-chemistry’ model, which includes stepwise and Penning ionization. Standing striations appeared in Townsend and glow discharges at low currents, and moving striations were obtained for the discharge currents exceeding a critical value. These waves originate at the anode and propagate towards the cathode. We have seen two types of moving striations with the two-level and full-chemistry models, which resemble the s and p striations previously observed in the experiments. Simulations indicate that processes in the anode region could control moving striations in the positive column plasma. The developed model helps clarify the nature of standing and moving striations in DC discharges of noble gases at low discharge currents and low gas pressures.

Physics↗

Exactly solvable relativistic model with the anomalous interaction

A special class of Dirac-Pauli equations with time-like vector potentials of an external field is investigated. An exactly solvable relativistic model describing the anomalous interaction of a neutral Dirac fermion with a cylindrically symmetric external electromagnetic field is presented. The related external field is a superposition of the electric field generated by a charged infinite filament and the magnetic field generated by a straight line current. In the nonrelativistic approximation the considered model is reduced to the integrable Pron'ko-Stroganov model.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

The Simons Observatory: Combining cross-spectral foreground cleaning with multitracer B -mode delensing for improved constraints on inflation

The Simons Observatory (SO), due to start full science operations in early 2025, aims to set tight constraints on inflationary physics by inferring the tensor-to-scalar ratio r from measurements of cosmic microwave background (CMB) polarization B-modes. Its nominal design including three small-aperture telescopes (SATs) targets a precision σ⁡(r = 0) ≤ 0.003 without delensing. Achieving this goal and further reducing uncertainties requires a thorough understanding and mitigation of other large-scale B-mode sources such as Galactic foregrounds and weak gravitational lensing. We present an analysis pipeline aiming to estimate r by including delensing within a cross-spectral likelihood, and demonstrate it for the first time on SO-like simulations accounting for various levels of foreground complexity, inhomogeneous noise and partial sky coverage. As introduced in an earlier SO delensing paper, lensing B-modes are synthesized using internal CMB lensing reconstructions as well as Planck-like cosmic infrared background maps and LSST-like galaxy density maps. We then extend SO’s power-spectrum-based foreground-cleaning algorithm to include all auto- and cross-spectra between the lensing template and the SAT B-modes in the likelihood function. This allows us to constrain r and the parameters of our foreground model simultaneously. Within this framework, we demonstrate the equivalence of map-based and cross-spectral delensing and use it to motivate an optimized pixel-weighting scheme for power spectrum estimation. We start by validating our pipeline in the simplistic case of uniform foreground spectral energy distributions. In the absence of primordial B-modes, we find that the 1⁢σ statistical uncertainty on r, σ⁡(r), decreases by 37% as a result of delensing. Tensor modes at the level of r = 0.01 are successfully detected by our pipeline. Even when using more realistic foreground models including spatial variations in the dust and synchrotron spectral properties, we obtain unbiased estimates of r both with and without delensing by employing the moment-expansion method. In this case, uncertainties are increased due to the higher number of model parameters, and delensing-related improvements range between 27% and 31%. These results constitute the first realistic assessment of the delensing performance at SO’s nominal sensitivity level.

79 ASTRONOMY AND ASTROPHYSICS↗