Search NASA⌕ Search

SEARCH · Search NASA

Results for “differentiable programming”

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 235 records · Page 13
All sourcesNASA NTRSDOE OSTI
DOE OSTI2012

Photoactive transparent nano-crystalline glass-ceramic for remazole red dye degradation

Graphical abstract: Display Omitted Highlights: ► Preparation and characterization of novel transparent nanocrystalline glass-ceramic. ► Precipitation of photoactive phases by using controlled heat-treatment. ► Conservation of transparency along with photoactivity. ► Using the prepared nanocrystalline glass-ceramic in water purification. -- Abstract: Transparent glass ceramic material was prepared from alkali-borosilicate glass containing titania by proper heat treatment scheme. The prepared samples were characterized using differential thermal analysis, X-ray diffraction, transmission electron microscope, selected area electron diffraction and UV–visible spectroscopy. The applied heat treatment program allowed the crystallization of nano-crystalline anatase, rutile, barium titanate, titanium borate and silicate phases while maintaining the transparency. The precipitated nano-crystalline anatase and rutile phases were responsible for the observed high photocatalytic activity of the prepared samples. Samples of 24.29 and 32.39 TiO{sub 2} wt% showed better efficiency for the decolorization of remazole red dye compared with commercial-TiO{sub 2} used in preparation of glass-ceramic. The reuse of prepared glass-ceramic photocatalyst with nearly same efficiency for different times was also proved.

77 NANOSCIENCE AND NANOTECHNOLOGY↗
DOE OSTI2015

Comparison of Calibration of Sensors Used for the Quantification of Nuclear Energy Rate Deposition

This present work deals with a collaborative program called GAMMA-MAJOR 'Development and qualification of a deterministic scheme for the evaluation of GAMMA heating in MTR reactors with exploitation as example MARIA reactor and Jules Horowitz Reactor' between the National Centre for Nuclear Research of Poland, the French Atomic Energy and Alternative Energies Commission and Aix Marseille University. One of main objectives of this program is to optimize the nuclear heating quantification thanks to calculation validated from experimental measurements of radiation energy deposition carried out in irradiation reactors. The quantification of the nuclear heating is a key data especially for the thermal, mechanical design and sizing of irradiation experimental devices in specific irradiated conditions and locations. The determination of this data is usually performed by differential calorimeters and gamma thermometers such as used in the experimental multi-sensors device called CARMEN 'Calorimetric en Reacteur et Mesures des Emissions Nucleaires'. In the framework of the GAMMA-MAJOR program a new calorimeter was designed for the nuclear energy deposition quantification. It corresponds to a single-cell calorimeter and it is called KAROLINA. This calorimeter was recently tested during an irradiation campaign inside MARIA reactor in Poland. This new single-cell calorimeter differs from previous CALMOS or CARMEN type differential calorimeters according to three main points: its geometry, its preliminary out-of-pile calibration, and its in-pile measurement method. The differential calorimeter, which is made of two identical cells containing heaters, has a calibration method based on the use of steady thermal states reached by simulating the nuclear energy deposition into the calorimeter sample by Joule effect; whereas the single-cell calorimeter, which has no heater, is calibrated by using the transient thermal response of the sensor (heating and cooling steps). The paper will concern these two kinds of calorimetric sensors. It will focus in particular on studies on their out-of-pile calibrations. Firstly, the characteristics of the sensor designs will be detailed (such as geometry, dimension, material sample, assembly, instrumentation). Then the out-of-pile calibration methods will be described. Furthermore numerical results obtained thanks to 2D axisymmetrical thermal simulations (Finite Element Method, CAST3M) and experimental results will be presented for each sensor. A comparison of the two different thermal sensor behaviours will be realized. To conclude a discussion of the advantages and the drawbacks of each sensor will be performed especially regarding measurement methods. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2018

Necrostatin-7 suppresses RANK-NFATc1 signaling and attenuates macrophage to osteoclast differentiation

Highlights: • Nec-7, but not Nec-1, inhibited osteoclast formation and function. • Nec-7 did not affect the M-CSF-supported survival of osteoclast precursor cells. • Nec-7 selectively attenuated RANK-NFATc1 axis and osteoclastic gene expression. • Forced overexpression of RANK restored osteoclastogenesis and NFATc1 amplification. Osteoclasts play a crucial role in osteolytic bone diseases, such as osteoporosis, rheumatoid arthritis, periodontitis, Paget's disease of bone and bone metastatic tumors. Therefore, controlling osteoclast differentiation and function has been considered a promising therapeutic strategy. Here, we show that necrostatin (Nec)-7, an inhibitor of programmed necrosis, strongly suppressed receptor activator of nuclear factor (NF)-κB ligand (RANKL)-induced osteoclastogenesis and bone resorption, without compromising macrophage colony-stimulating factor (M-CSF)-supported survival and growth of osteoclast precursor cells. Accordingly, Nec-7 significantly decreased the levels of RANKL-induced osteoclastogenic marker genes, such as cathepsin K. Mechanistically, Nec-7 neither affected MAPK nor NF-κB activation; however, it strongly inhibited the RANKL receptor (RANK) to nuclear factor of activated T cells c1 (NFATc1) signaling. Lentiviral expression of RANK in bone marrow-derived macrophages significantly restored osteoclastogenesis and NFATc1 amplification in Nec-7-treated cells. In this study, we revealed that Nec-7-sensitive pathways are crucially involved in osteoclast formation and function. Investigation of the molecular mechanism(s) through which Nec-7 inhibits RANK-NFATc1 signaling axis may lead to the development of new therapeutic strategies for bone disease.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2014

Low-temperature random matrix theory at the soft edge

“Low temperature” random matrix theory is the study of random eigenvalues as energy is removed. In standard notation, β is identified with inverse temperature, and low temperatures are achieved through the limit β → ∞. In this paper, we derive statistics for low-temperature random matrices at the “soft edge,” which describes the extreme eigenvalues for many random matrix distributions. Specifically, new asymptotics are found for the expected value and standard deviation of the general-β Tracy-Widom distribution. The new techniques utilize beta ensembles, stochastic differential operators, and Riccati diffusions. The asymptotics fit known high-temperature statistics curiously well and contribute to the larger program of general-β random matrix theory.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2012

Development of a commercial enzymes system for lignocellulosic biomass saccharification

DSM Innovation Inc., in its four year effort was able to evaluate and develop its in-house DSM fungal cellulolytic enzymes system to reach enzyme efficiency mandates set by DoE Biomass program MYPP goals. DSM enzyme cocktail is uniquely active at high temperature and acidic pH, offering many benefits and product differentiation in 2G bioethanol production. Under this project, strain and process development, ratio optimization of enzymes, protein and genetic engineering has led to multitudes of improvement in productivity and efficiency making development of a commercial enzyme system for lignocellulosic biomass saccharification viable. DSM is continuing further improvement by additional biodiversity screening, protein engineering and overexpression of enzymes to continue to further lower the cost of enzymes for saccharification of biomass.

09 BIOMASS FUELS↗
DOE OSTI2020

Coupled Time-Lapse Full-Waveform Inversion for Subsurface Flow Problems Using Intrusive Automatic Differentiation

We describe a novel framework for estimating subsurface properties, such as rock permeability and porosity, from time-lapse observed seismic data by coupling full-waveform inversion (FWI), subsurface flow processes, and rock physics models. For the inverse modeling, we handle the back propagation of gradients by an intrusive automatic differentiation strategy that offers three levels of user control: (1) At the wave physics level, we adopted the discrete adjoint method in order to use our existing high-performance FWI code; (2) at the rock physics level, we used built-in automatic differentiation operators from the TensorFlow backend; (3) at the flow physics level, we implemented customized partial differential equation (PDE) operators for the multiphase flow equations. The three-level coupled inversion strategy strikes a good balance between computational efficiency and programming efforts, and when the gradients are chained together, it constitutes a coupled inverse system. Our numerical experiments demonstrate that the three-level coupled inverse problem is superior in terms of accuracy to a traditional decoupled inversion strategy. Additionally, our method is able to simultaneously invert for parameters in empirical relationships such as the rock physics models. Our proposed inverted model can be used for reservoir performance prediction and reservoir management/optimization purposes.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

CHARGED-CURRENT DOUBLE-DIFFERENTIAL CROSS SECTION MEASUREMENT OF ELECTRON NEUTRINOS IN LOW RECOIL REGION ON HYDROCARBON AT ∼7 GEV

Measuring neutrino mixing angles, especially the CP violation phase and mass hierarchy, is one of the priorities of the neutrino community and long baseline neutrino oscillation programs, such as T2K and NOvA, provided the most precise measurements to date. One of the dominating uncertainties of the measurements is the neutrino interaction model due to the difficulty of modeling the nuclear effects from first principles. As a result, neutrino scattering experiments such as MINERvA were conducted to provide data as input to reduce this uncertainty. This thesis presents a charged-current double-differential cross section measurement of electron neutrinos using the MINERvA detector in terms of available energy (a proxy of energy transfer), three-momentum transfer, and/or transverse momentum of the final state electron. This measurement focuses on the transition region between quasi-elastic and resonance processes, where the multi-nucleon knockout process dominates. The result is compared to a modified GENIE model and similar measurements using muon neutrino samples. In addition, this measurement reproduces the photon-like excess seen by previous MINERvA electron neutrino measurements.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2022

Exponential Decay of Sensitivity in Graph-Structured Nonlinear Programs

We study solution sensitivity for nonlinear programs (NLPs) whose structures are induced by graphs. These NLPs arise in many applications such as dynamic optimization, stochastic optimization, optimization with partial differential equations, and network optimization. We illustrate that for a given pair of nodes, the sensitivity of the primal-dual solution at one node against a data perturbation at the other node decays exponentially with respect to the distance between these two nodes on the graph. In other words, the solution sensitivity decays as one moves away from the perturbation point. This result, which we call exponential decay of sensitivity, holds under the strong second-order sufficiency condition and the linear independence constraint qualification. We also present conditions under which the decay rate remains uniformly bounded; this allows us to characterize the sensitivity behavior of NLPs defined over subgraphs of infinite graphs. The theoretical developments are illustrated with numerical examples.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2016

Nucleon resonance structure studies via exclusive KY electroproduction

Studying the structure of excited nucleon states employing the electroproduction of exclusive reactions is an important avenue for exploring the nature of the non-perturbative strong interaction. The electrocouplings of $N^*$ states in the mass range below 1.8~GeV have been determined from analyses of CLAS $\pi N$, $\eta N$, and $\pi \pi N$ data. This work has made it clear that consistent results from independent analyses of several exclusive channels with different couplings and non-resonant backgrounds but the same $N^*$ electro excitation amplitudes, is essential to have confidence in the extracted results. In terms of hadronic coupling, many high-lying $N^*$ states preferentially decay through the $\pi \pi N$ channel instead of $\pi N$. Data from the $KY$ channels will therefore be critical to provide an independent analysis to compare the extracted electrocouplings for the high-lying $N^*$ states against those determined from the $\pi N$ and $\pi \pi N$ channels. Lastly, a program to study excited $N^*$ state structure in both non-strange and strange exclusive electroproduction channels using CLAS12 will measure differential cross sections and polarization observables to be used as input to extract the $\gamma_vNN^*$ electrocoupling amplitudes for the most prominent $N^*$ states in the range of invariant energy $W$ up 3~GeV in the virtually unexplored domain of momentum transfers $Q^2$ up to 12~GeV$^2$.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2022

PythonFOAM: In-situ data analyses with OpenFOAM and Python

Here, we outline the development of a general-purpose Python-based data analysis tool for OpenFOAM. Our implementation relies on the construction of OpenFOAM applications that have bindings to data analysis libraries in Python. Double precision data in OpenFOAM is cast to a NumPy array using the NumPy C-API and Python modules may then be used for arbitrary data analysis and manipulation on flow-field information. We highlight how the proposed wrapper may be used for an in-situ online singular value decomposition (SVD) implemented in Python and accessed from the OpenFOAM solver PimpleFOAM. Here, 'in-situ' refers to a programming paradigm that allows for a concurrent computation of the data analysis on the same computational resources utilized for the partial differential equation solver. In addition, to demonstrate parallel deployments, we deploy a distributed SVD, which collects snapshot data across the ranks of a distributed simulation to compute the global left singular vectors. Crucially, both OpenFOAM and Python share the same message passing interface (MPI) communicator for this deployment which allows Python objects and functions to exchange NumPy arrays across ranks. Subsequently, we provide scaling assessments of this distributed SVD on multiple nodes of Intel Broadwell and KNL architectures for canonical test cases such as the large eddy simulations of a backward facing step and a channel flow at friction Reynolds number of 395. Finally, we demonstrate the deployment of a deep neural network for compressing the flow-field information using an autoencoder to demonstrate an ability to use state-of-the-art machine learning tools in the Python ecosystem.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Energy Dependent Fission Product Yields (2nd year Milestone and Deliverables)

Introduction to the project: One of the Laboratory missions is to provide consistent, high-precision fission product; yield data critical for testing fission models and maintaining the safety and security of the nation’s nuclear weapons stockpile. However, high-quality, energy-differential fission product yield data is missing for certain actinides and neutron energies important to constrain the new U.S. Nuclear Data Program evaluation effort; It has been shown that the reactor antineutrino anomaly may be at least partially caused by roughly 20 fission products. The fission product yield data is missing or incomplete for many of these isotopes, thus it is necessary to accurately determine these values to better constrain the anomaly; Aitor Bracho is measuring very short-lived (seconds to minutes) fission product yields of 235 U and 239 Pu using monoenergetic neutron beams at E n = 60 and 560 keV; Aitor Bracho is using a direct approach utilizing a state-of-the-art rabbit transfer system, superior HPGe detector, and digital acquisition systems for fission decay measurements. The goals of this project: lop experimental capabilities and data analysis techniques to carry out the gamma-ray spectra analysis necessary for fission product yield calculation; Provide high-precision and energy-dependent fission product data supporting fission theory, neutrino physics, and applied physics.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2017

PDE-constrained Optimization under Uncertainty

Uncertainty is pervasive in all science and engineering applications. Incorporating uncertainty in physical models is therefore both natural and vital. In doing so, we often arrive at parametric systems of partial differential equations (PDEs). When passing from simulation to optimization, we obtain (typically nonconvex) infinite-dimensional optimization problems that, upon discretization, result in extremely large-scale nonlinear programs. For example, consider a linear elliptic PDE on a two- dimensional domain with a single random coeffcient. If we sampled the random input with 10,000 realizations of the coeffcient, the resulting optimization problem would have 10,000 PDE constraints. Furthermore, discretizing each PDE with piecewise linear finite elements on a 100X100 uni- form quadrilateral mesh results in 100,000,000 degrees of freedom. As a result, the critical components for ensuring mesh-independent performance of numerical optimization methods in the deterministic setting, for example, solution regularity and generalized differentiability, are even more critical in the stochastic setting.

42 ENGINEERING↗
DOE OSTI2016

Development of Techniques for Spent Fuel Assay – Differential Dieaway Final Report

This report summarizes the work done under a DNDO R&D funded project on the development of the differential dieaway method to measure plutonium in spent fuel. There are large amounts of plutonium that are contained in spent fuel assemblies, and currently there is no way to make quantitative non-destructive assay. This has led NA24 under the Next Generation Safeguards Initiative (NGSI) to establish a multi-year program to investigate, develop and implement measurement techniques for spent fuel. The techniques which are being experimentally tested by the existing NGSI project do not include any pulsed neutron active techniques. The present work covers the active neutron differential dieaway technique and has advanced the state of knowledge of this technique as well as produced a design for a practical active neutron interrogation instrument for spent fuel. Monte Carlo results from the NGSI effort show that much higher accuracy (1-2%) for the Pu content in spent fuel assemblies can be obtained with active neutron interrogation techniques than passive techniques, and this would allow their use for nuclear material accountancy independently of any information from the operator. The main purpose of this work was to develop an active neutron interrogation technique for spent nuclear fuel.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2025

Parallel sorting algorithm classification: is manual instrumentation necessary?

Understanding parallel algorithms is crucial for accelerating scientific simulations on complex, distributed memory, high-performance computers. Modern algorithm classification approaches learn semantics directly from source code to differentiate between algorithms, however, accessing source code is not always possible. We can learn about parallel algorithms from observing their performance, as programs running the same algorithms and using the same hardware should exhibit similar performance characteristics. We present an approach to learn algorithm classes from parallel performance data directly in order to classify algorithms without access to the source code. We extend previous work to enable classifying parallel sorting algorithms using automatic instrumentation instead of requiring manual region annotations in the source code. In this work, we design and demonstrate a study for classification of parallel sorting algorithms using parallel performance data collected from automatic instrumentation, and evaluate the performance of our new methodology on classification. We leverage Caliper to collect the performance data, Thicket for our exploratory data analysis (EDA), and PyTorch and Scikit-learn to evaluate the effectiveness of random forests, support vector machines (SVMs), decision trees, neural networks, and logistic regressions on parallel performance data. Additionally, we study noise in parallel performance data, whether the removal of noise and pre-processing of the data is necessary to accurately classify parallel sorting algorithms, and determine the effectiveness of features created from performance data. In conclusion, we demonstrate classification accuracy for these five different models of up to 97.7% across four different parallel algorithm classes.

Algorithm Classification↗
DOE OSTI2025

Adaptive PID Gain Scheduling Control for Hydropower Turbine Using Neural CDE and Stochastic Distribution Shaping

This paper introduces a gain-scheduling PID controller design strategy for hydroturbine frequency control mode. This scheme first uses real data to learn the nonlinear dynamics of the hydroturbine using neural controlled differential equations and then perturbs the obtained nonlinear system at different equilibrium points, based on which a static output feedback adaptive dynamic programming algorithm is then used to optimize the PID gains for each equilibrium point. Moreover, a continuous-time version of stochastic distribution control is proposed to further fine-tune the optimized PID gains. Finally, the controller is obtained by implementing linear interpolation between the optimized PID control gains. The simulation results show that the proposed gain-scheduling PID controller can control a larger range of operation points compared with the given fixed PID controller and the baseline method. Compared with the given fixed PID controller, the proposed gain-scheduling PID controller can regulate hydroturbine frequency against disturbances induced by power-load variation with over 50% less overshoot for some operation points.

13 HYDRO ENERGY↗
DOE OSTI2021

Example on how to (intelligently) augment the nuclear-data pipeline with machine learning [Slides]

The presentation discusses how machine learning has helped the Los Alamos National Laboratory (LANL) nuclear-data pipeline. It also discusses the strengths of machine learning as it finds trends in large amounts of data where human brains are overwhelmed and that this information may be crucial to improve our nuclear data. It does stress, however, that machine learning is no "silver bullet" and that it is critical to feed it expert knowledge and use physics intuition to interpret the results. The presentation discusses the need to develop infrastructure and tools to provide data in an easily readable and unambiguously interpretable format (e.g., EXFOR format), to develop experimental data and theory to solve physics questions, and that statisticians and nuclear-data experts must be brought together to correctly interpret the results. The presentation concludes by stating that machine learning is a great tool and that LANL needs to use the algorithms along with developing physics data, tools and infrastructure.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2012

Investigation of near-threshold eta-meson production in the reaction {pi}{sup -}p{yields} {eta}n

Differential and total cross sections for eta-meson production in the reaction {pi}{sup -}p {yields} {eta}n were measured within the experimental program eta-meson physics implemented in the pion channel of the synchrocyclotron of the Petersburg Nuclear Physics Institute (PNPI, Gatchina). These measurements were performed at incident-pion momenta (700, 710, 720, and 730 MeV/c) in the vicinity of the threshold for the process under study by using the neutral-meson spectrometer designed and created at the Meson Physics Laboratory of PNPI. It is shown that, in the immediate vicinity of the threshold (685 MeV/c), the process of eta-meson production proceeds predominantly via S{sub 11}(1535)-resonance formation followed by the decay S{sub 11}(1535) {yields} {eta}n (the respective branching fraction is Br Almost-Equal-To 60%), but that, as the momentum of incident pions increases, the role of the D wave becomes ever more important. A detailed analysis of this effect indicates that it is due to the increasing contribution of the D{sub 13}(1520) resonance. Although the branching fraction of the decay of this resonance through the {eta}n channel is assumed to be very small (BR Almost-Equal-To 0.24%), the effect is enhanced owing to the interference between the D wave and the dominant resonance S{sub 11}(1535).

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗