Search NASA⌕ Search

SEARCH · Search NASA

Results for “embedding model”

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 91 records · Page 5

TUNABLE EOS MATERIAL MODEL IN THE SIMULATION OF PULSED MERCURY SPALLATION TARGET VESSEL

A pulsed spallation target is subjected to very short (~1μs) but intense loads from repeated proton pulses. The effect of this pulsed loading on the stainless-steel target module which contains the flowing mercury target material is difficult to predict. Different simulation approaches and material models for the mercury have been tried. To date the best match of simulation results to experimental data was obtained with an equation of state (EOS) material model with a specified tensile cutoff pressure which simulates the cavitation threshold [1]. The inclusion of a threshold to represent cavitation was a key parameter in achieving successful predictions of stress waves triggered by the high energy pulse striking the mercury and vessel. However, recent measurements of strain responses of target modules show that significant differences between the measured strain and the strain simulated with EOS mercury model still exist. These differences grow to irreconcilable values when non-condensable helium gas is intentionally injected into the flowing mercury. A novel EOS mercury model embedded into ABAQUS VUMAT has been investigated in this project, which introduces the concept of proportional, integral, and derivative (PID) control into the mercury EOS model. By tuning the new introduced PID parameters (Kp, Ki and Kd), we replace the specified cutoff pressure with an adjustable spring-damper-like material behavior which may better match the complex dynamics of the mercury and helium mixture. This approach is expected to reduce the gap between measured and simulated vessel strain responses. Primitive application of this tunable EOS mercury model on prototypic shape experimental target has shown its capability and potentiality of improving mechanical behaviors of EOS mercury with cutoff pressure considered.

Lin, Lianshan↗

Processing MCNP Elemental Edit Outputs

The Monte Carlo N-Particle (MCNP) transport code version 6 (also known as MCNP6) has the capability for tracking particles on unstructured mesh (UM) geometry models embedded into constructive solid geometry (CSG) cells. A UM geometry is a collection of elements representing a solid geometry. The first step of MCNP UM modeling is using other software packages to create a finite element mesh representation of a solid 3D geometry. Computer-aided design (CAD) or computer-aided manufacturing (CAM) software is typically used to create a solid geometry model, which is later imported into mesh generation software to create a UM model. The MCNP UM feature was originally designed for models generated by the Abaqus/CAE software. The MCNP code version 6.0 and later can process UM models formatted as Abaqus input files. MCNP can process a UM model consisting of several different element types including linear tetrahedral or hexahedral elements and calculate quantities of interest such as flux and energy deposition at elements. An MCNP UM simulation provides high-fidelity elemental edit (i.e., tally) outputs, which can be further used in multiphysics calculations. The MCNP UM feature was used for multiphysics simulations where quantities of interest calculated by MCNP are used as inputs for heat transfer calculations in Abaqus. MCNP6.3 can produce two types of elemental edit output (EEOUT) file formats: ASCII and HDF5. An EEOUT file type must be requested on an EMBED card while output type (flux or energy deposition) must be requested on an EMBEE card. We wrote Python3 scripts to extract energy deposition values in an ASCII or HDF5 EEOUT file and compute a heat flux profile for an Abaqus heat transfer calculation.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

An Embedded 3D Fracture Modeling Approach for Simulating Fracture-Dominated Fluid Flow and Heat Transfer in Geothermal Reservoirs

An efficient modeling approach is described for incorporating arbitrary 3D, discrete fractures, such as hydraulic fractures or faults, into modeling fracture-dominated fluid flow and heat transfer in fractured geothermal reservoirs. This technique allows 3D discrete fractures to be discretized independently from surrounding rock volume and inserted explicitly into a primary fracture/matrix grid, generated without including 3D discrete fractures in prior. An effective computational algorithm is developed to discretize these 3D discrete fractures and construct local connections between 3D fractures and fracture/matrix grid blocks of representing the surrounding rock volume. The constructed gridding information on 3D fractures is then added to the primary grid. This embedded fracture modeling approach can be directly implemented into a developed geothermal reservoir simulator via the integral finite difference (IFD) method or with TOUGH2 technology This embedded fracture modeling approach is very promising and computationally efficient to handle realistic 3D discrete fractures with complicated geometries, connections, and spatial distributions. Compared with other fracture modeling approaches, it avoids cumbersome 3D unstructured, local refining procedures, and increases computational efficiency by simplifying Jacobian matrix size and sparsity, while keeps sufficient accuracy. Several numeral simulations are present to demonstrate the utility and robustness of the proposed technique. Our numerical experiments show that this approach captures all the key patterns about fluid flow and heat transfer dominated by fractures in these cases. Thus, this approach is readily available to simulation of fractured geothermal reservoirs with both artificial and natural fractures.

15 GEOTHERMAL ENERGY↗

Creating Training Data for Scientific Named Entity Recognition with Minimal Human Effort

Scientific Named Entity Referent Extraction is often more complicated than traditional Named Entity Recognition (NER). For example, in polymer science, chemical structure may be encoded in a variety of nonstandard naming conventions, and authors may refer to polymers with conventional names, commonly used names, labels (in lieu of longer names), synonyms, and acronyms. As a result, accurate scientific NER methods are often based on task-specific rules, which are difficult to develop and maintain, and are not easily generalized to other tasks and fields. Machine learning models require substantial expert-annotated data for training. Here we propose polyNER: a semi-automated system for efficient identification of scientific entities in text. PolyNER applies word embedding models to generate entity-rich corpora for productive expert labeling, and then uses the resulting labeled data to bootstrap a context-based word vector classifier. Evaluation on materials science publications shows that the polyNER approach enables improved precision or recall relative to a state-of-the-art chemical entity extraction system at a dramatically lower cost: it required just two hours of expert time, rather than extensive and expensive rule engineering, to achieve that result. This result highlights the potential for human-computer partnership for constructing domain-specific scientific NER systems.

Crowdsourcing↗

Experimental verification of a crystal plasticity-based simulation framework for predicting microstructure and geometric shape changes: Application to bending and Taylor impact testing of Zr

This paper is concerned with experimental verification of a recently developed multi-scale simulation framework for plastic deformation of metallic materials from quasi-static to impact deformation conditions. The framework is a visco-plastic self-consistent (VPSC) polycrystalline model embedded in an implicit finite element method (FE-VPSC) to provide a microstructure-sensitive constitutive response at each material point. Each material point of the FEM model is a polycrystalline aggregate with crystallographic deformation mechanisms operating at the single crystal scale with their evolving activity based on a dislocation density-based hardening law and texture. Four beams and three cylinders machined in different orientations from a textured plate of high-purity zirconium are tested quasi-statically in 4-point bending and at speeds of 100 m/s, 170 m/s and 243 m/s during Taylor impact tests, respectively. The variation in dimensional changes resulting from different sample orientations in the plate with respect to loading directions is measured for each sample. Moreover, texture and twinning characterization is performed using electron backscattered diffraction (EBSD). The deformation processes and underlying evolution of microstructure are successfully simulated using the FE-VPSC framework. In doing so, the model parameters are optimized and validated across a broad range of strain rates and temperatures. Simulation results in terms of geometrical changes and microstructural evolution are compared with the experimental measurements. The model predicts anisotropic material flow resulting from the hard-to-deform crystallographic directions, the development of gradients in texture and twinning through the geometries, tension–compression asymmetry, as well as the extent of plasticity under impact.

42 ENGINEERING↗

$\mathrm{PPT}$-Multicore: performance prediction of Open$\mathrm{MP}$ applications using reuse profiles and analytical modeling

In this report we present PPT-Multicore, an analytical model embedded in the Performance Prediction Toolkit (PPT) to predict parallel applications’ performance running on a multicore processor. PPT-Multicore builds upon our previous work towards a multicore cache model. We extract LLVM basic block labeled memory trace using an architecture-independent LLVM-based instrumentation tool only once in an application’s lifetime. The model uses the memory trace and other parameters from an instrumented sequentially executed binary. We use probabilistic and computationally efficient reuse profiles to predict the cache hit rates and runtimes of OpenMP programs’ parallel sections. We model Intel’s Broadwell, Haswell, and AMD’s Zen2 architectures and validate our framework using different applications from PolyBench and PARSEC benchmark suites. The results show that PPT-Multicore can predict cache hit rates with an overall average error rate of 1.23% while predicting the runtime with an error rate of 9.08%.

97 MATHEMATICS AND COMPUTING↗

Calibrating hypersonic turbulence flow models with the HIFiRE-1 experiment using data-driven machine-learned models.

In this paper we study the efficacy of combining machine-learning methods with projection-based model reduction techniques for creating data-driven surrogate models of computationally expensive, high-fidelity physics models. Such surrogate models are essential for many-query applications e.g., engineering design optimization and parameter estimation, where it is necessary to invoke the high-fidelity model sequentially, many times. Surrogate models are usually constructed for individual scalar quantities. However there are scenarios where a spatially varying field needs to be modeled as a function of the model’s input parameters. Here we develop a method to do so, using projections to represent spatial variability while a machine-learned model captures the dependence of the model’s response on the inputs. The method is demonstrated on modeling the heat flux and pressure on the surface of the HIFiRE-1 geometry in a Mach 7.16 turbulent flow. The surrogate model is then used to perform Bayesian estimation of freestream conditions and parameters of the SST (Shear Stress Transport) turbulence model embedded in the high-fidelity (Reynolds-Averaged Navier–Stokes) flow simulator, using shock-tunnel data. The paper provides the first-ever Bayesian calibration of a turbulence model for complex hypersonic turbulent flows. We find that the primary issues in estimating the SST model parameters are the limited information content of the heat flux and pressure measurements and the large model-form error encountered in a certain part of the flow.

42 ENGINEERING↗

SpeckleNN: a unified embedding for real-time speckle pattern classification in X-ray single-particle imaging with limited labeled examples

With X-ray free-electron lasers (XFELs), it is possible to determine the three-dimensional structure of noncrystalline nanoscale particles using X-ray single-particle imaging (SPI) techniques at room temperature. Classifying SPI scattering patterns, or `speckles', to extract single-hits that are needed for real-time vetoing and three-dimensional reconstruction poses a challenge for high-data-rate facilities like the European XFEL and LCLS-II-HE. Here, we introduce SpeckleNN, a unified embedding model for real-time speckle pattern classification with limited labeled examples that can scale linearly with dataset size. Trained with twin neural networks, SpeckleNN maps speckle patterns to a unified embedding vector space, where similarity is measured by Euclidean distance. We highlight its few-shot classification capability on new never-seen samples and its robust performance despite having only tens of labels per classification category even in the presence of substantial missing detector areas. Without the need for excessive manual labeling or even a full detector image, our classification method offers a great solution for real-time high-throughput SPI experiments.

47 OTHER INSTRUMENTATION↗

TransformerG2G: Adaptive time-stepping for learning temporal graph embeddings using transformers

Dynamic graph embedding has emerged as a very effective technique for addressing diverse temporal graph analytic tasks (i.e., link prediction, node classification, recommender systems, anomaly detection, and graph generation) in various applications. Such temporal graphs exhibit heterogeneous transient dynamics, varying time intervals, and highly evolving node features throughout their evolution. Hence, incorporating long-range dependencies from the historical graph context plays a crucial role in accurately learning their temporal dynamics. In this paper, we develop a graph embedding model with uncertainty quantification, TransformerG2G, by exploiting the advanced transformer encoder to first learn intermediate node representations from its current state (t) and previous context (over timestamps [t–1,t–l], l is the length of context). Moreover, we employ two projection layers to generate lower-dimensional multivariate Gaussian distributions as each node's latent embedding at timestamp t. We consider diverse benchmarks with varying levels of "novelty" as measured by the TEA (Temporal Edge Appearance) plots. Here, our experiments demonstrate that the proposed TransformerG2G model outperforms conventional multi-step methods and our prior work (DynG2G) in terms of both link prediction accuracy and computational efficiency, especially for high degree of novelty. Furthermore, the learned time-dependent attention weights across multiple graph snapshots reveal the development of an automatic adaptive time stepping enabled by the transformer. Importantly, by examining the attention weights, we can uncover temporal dependencies, identify influential elements, and gain insights into the complex interactions within the graph structure. For example, we identified a strong correlation between attention weights and node degree at the various stages of the graph topology evolution.

97 MATHEMATICS AND COMPUTING↗

The age–chemical abundance structure of the Galactic disc – II. α -dichotomy and thick disc formation

ABSTRACT We extend our previous work on the age–chemical abundance structure of the Galactic outer disc to the inner disc (4 < r < 8 kpc) based on the SDSS/APOGEE survey. Different from the outer disc, the inner disc stars exhibit a clear bimodal distribution in the [Mg/Fe]–[Fe/H] plane. While a number of scenarios have been proposed in the literature, it remains challenging to recover this bimodal distribution with theoretical models. To this end, we present a chemical evolution model embedding a complex multiphase inner disc formation scenario that matches the observed bimodal [Mg/Fe]–[Fe/H] distribution. In this scenario, the formation of the inner disc is dominated by two main starburst episodes $6\,$Gyr apart with secular, low-level star formation activity in between. In our model, the first starburst occurs at early cosmic times ($t\sim 1\,$ Gyr) and the second one $6\,$ Gyr later at a cosmic time of $t\sim 7\,$ Gyr. Both these starburst episodes are associated with gas accretion events in our model, and are quenched rapidly. The first starburst leads to the formation of the high-α sequence, and the second starburst leads to the formation of the metal-poor low-α sequence. The metal-rich low-α stars, instead, form during the secular evolution phase between the two bursts. Our model shows that the α-dichotomy originates from the rapid suppression of star formation after the first starburst. The two starburst episodes are likely to be responsible for the formation of the geometric thick disc (z >1 kpc), with the old inner thick disc and the young outer thick disc forming during the first and the second starbursts, respectively.

Lian, Jianhui↗

REFLECTED CHARGED PARTICLE POPULATIONS AROUND DIPOLAR LUNAR MAGNETIC ANOMALIES

In this work we analyze and compare the reflected particle populations for both a horizontal and a vertical dipole model embedded in the lunar surface, representing the solar wind interaction with two different lunar magnetic anomaly (LMA) structures. Using the 3D full-kinetic electromagnetic code iPic3D, in combination with a test-particle approach to generate particle trajectories, we focus on the ion and electron dynamics. Whereas the vertical model electrostatically reflects ions upward under both near-parallel and near-perpendicular angles with respect to the lunar surface, the horizontal model only has a significant shallow component. Characterizing the electron dynamics, we find that the interplay of the mini-magnetosphere electric and magnetic fields is capable of temporarily trapping low-energy electrons and possibly ejecting them upstream. Our results are in agreement with recent high-resolution observations. Low- to medium-altitude ion and electron observations might be excellent indicators to complement orbital magnetic field measurements and better uncover the underlying magnetic field structure. The latter is of particular importance in defining the correlation between LMAs and lunar swirls, and further testing the solar wind shielding hypothesis for albedo markings due to space weathering. Observing more reflected ions does not necessarily point to the existence of a mini-magnetosphere.

79 ASTRONOMY AND ASTROPHYSICS↗

DeePMD-kit v2: A software package for deep potential models

DeePMD-kit is a powerful open-source software package that facilitates molecular dynamics simulations using machine learning potentials known as Deep Potential (DP) models. This package, which was released in 2017, has been widely used in the fields of physics, chemistry, biology, and material science for studying atomistic systems. The current version of DeePMD-kit offers numerous advanced features, such as DeepPot-SE, attention-based and hybrid descriptors, the ability to fit tensile properties, type embedding, model deviation, DP-range correction, DP long range, graphics processing unit support for customized operators, model compression, non-von Neumann molecular dynamics, and improved usability, including documentation, compiled binary packages, graphical user interfaces, and application programming interfaces. This article presents an overview of the current major version of the DeePMD-kit package, highlighting its features and technical details. Additionally, this article presents a comprehensive procedure for conducting molecular dynamics as a representative application, benchmarks the accuracy and efficiency of different models, and discusses ongoing developments.

97 MATHEMATICS AND COMPUTING↗

Brane world in non-Riemannian geometry

We carefully investigate the modified Einstein's field equation in a 4-dimensional (3-brane) arbitrary manifold embedded in a 5-dimensional non-Riemannian bulk spacetime with a noncompact extra dimension. In this context the Israel-Darmois matching conditions are extended assuming that the torsion in the bulk is continuous. The discontinuity in the torsion first derivatives are related to the matter distribution through the field equation. In addition, we develop a model that describes a flat FLRW model embedded in a 5-dimensional de Sitter or anti-de Sitter, where a 5-dimensional cosmological constant emerges from the torsion.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

A compact 341 model at TeV scale

We build a gauge model based on the SU(3){sub c}⊗SU(4){sub L}⊗U(1){sub X} symmetry where the scalar spectrum needed to generate gauge boson and fermion masses has a smaller scalar content than usually assumed in literature. We compute the running of its abelian gauge coupling and show that a Landau pole shows up at the TeV scale, a fact that we use to consistently implement those fermion masses that are not generated by Yukawa interactions, including neutrino masses. This is appropriately achieved by non renormalizable effective operators, suppressed by the Landau pole scale. Also, SU(3){sub c}⊗SU(3){sub L}⊗U(1){sub N} models embedded in this gauge structure are bound to be strongly coupled at this same energy scale, contrary to what is generally believed, and neutrino mass generation is rather explained through the same effective operators used in the larger gauge group. Besides, their nice features, as the existence of cold dark matter candidates and the ability to reproduce the observed standard model Higgs-like phenomenology, are automatically inherited by our model. Finally, our results imply that this model is constrained to be observed or discarded soon, since it must be realized at the currently probed energy scale in LHC. - Highlights: • We build a SU(4){sub L}⊗U(1){sub N} electroweak model with a reduced scalar spectrum. • The existence of a Landau pole at TeV scale is investigated. • Fermion masses in this model, including neutrinos, emerge from effective operators. • Imposition of an extra Z{sub 2} symmetry avoids fast proton decay.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

SU-E-T-527: Prior Knowledge Guided TomoTherapy Treatment Planning

Purpose: The quality and efficiency of radiotherapy treatment planning are highly planer dependent. Previously we have developed a statistical model to correlate anatomical features with dosimetry features of head and neck Tomotherapy treatment. The model enables us to predict the best achievable dosimetry for individual patient prior to treatment planning. The purpose of this work is to study if the prediction model can facilitate the treatment planning in both the efficiency and dosimetric quality. Methods: The anatomy-dosimetry correlation model was used to calculate the expected DVH for nine patients formerly treated. In Group A (3 patients), the model prediction agreed with the clinic plan; in Group B (3 patients), the model predicted lower larynx mean dose than the clinic plan; in Group C (3 patients), the model suggested the brainstem could be further spared. Guided by the prior knowledge, we re-planned all 9 cases. The number of interactions during the optimization process and dosimetric endpoints between the original clinical plan and model-guided re-plan were compared. Results: For Group A, the difference of target coverage and organs-at-risk sparing is insignificant (p>0.05) between the replan and the clinical plan. For Group B, the clinical plan larynx median dose is 49.4±4.7 Gy, while the prediction suggesting 40.0±6.2 Gy (p<0.05). The re-plan achieved 41.5±6.6 Gy, with similar dose on other structures as clinical plan. For Group C, the clinical plan brainstem maximum dose is 44.7±5.5 Gy. The model predicted lower value 32.2±3.8 Gy (p<0.05). The re-plans reduced brainstem maximum dose to 31.8±4.1 Gy without affecting the dosimetry of other structures. In the replanning of the 9 cases, the times operator interacted with TPS are reduced on average about 50% compared to the clinical plan. Conclusion: We have demonstrated that the prior expert knowledge embedded model improved the efficiency and quality of Tomotherapy treatment planning.

60 APPLIED LIFE SCIENCES↗

Reactor transient analyses with KIN3D/PARTISN

Efforts are going on at the Karlsruhe Institute of Technology (KIT) to extend the kinetics capability of the PARTISN code in order to run in parallel two- and three-dimensional transient analyses with the quasistatic method, while taking into account delayed neutrons. In the original code version, time-dependent transport problems are solved by employing a semi-implicit direct kinetics option, the delayed neutrons being not taken into account. The PARTISN 5.97 code has been extended and then coupled with KIN3D, a time-dependent model embedded in the ERANOS code system. In the coupled code, PARTISN 5.97 is used as neutron transport solver to perform transient analyses while employing direct and quasi-static kinetics options of KIN3D. The coupled code can be also applied for first-order and exact perturbation theory calculations. In the paper, the PARTISN 5.97 extensions and coupling procedure are described and the performances of the KIN3D/PARTISN coupled code are investigated by analyzing transients induced by a source-jerk in a three-dimensional ADS model driven by an external source. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Integrate FARM with PID controllers: IES Simulation Ecosystem Control System Development

The FARM (Feasible Actuator Range Modifier) software module is a component of the RAVEN-based FORCE framework for analysis of Integrated Energy Systems (IES). FARM was designed to support the HERON software module in the solution of the optimal dispatch problem for IES units. As the result of HERON-FARM dispatch simulation, the set-point trajectories are optimized to meet constraints on both the production variables (i.e., the variables to be optimized such as the electrical power, the hydrogen production rate, etc.) and the process variables tied to the service life of equipment (e.g., steam flowrate, vessel pressure, turbine firing temperature, etc.) at a coarse time resolution (every 10 or 100 seconds) over long time horizons (several days or weeks). In case the operational constraints need to be met at finer time resolution, the computational burden of HERON-FARM would linearly increase with the sampling rate, and sub-optimal solutions might be obtained. System responses characterized by overshoots and damped oscillations temporarily violating the imposed constraints might occur during abrupt power transients. In this report, a hierarchical control system architecture for the operation of the Dynamic Energy Technology and Integration Laboratory (DETAIL) facility constructed at INL was proposed. First, the preliminary studies on the proposed control strategy for operating the facility and the designed PI controllers were reviewed. In particular, the current approach for generating the set-point trajectories was studied, and its limits were identified. To this aim, the inclusion of a Supervisory Control layer embedding a modified version of the FARM algorithm for preserving the system safe operation over both long and real-time horizons was proposed. In this way, FARM would be applied twice, i.e., the original version (“FARM Validator”) aiding the solution of the power dispatch problem, and the modified version (“FARM Supervisory” coordinating the PI controllers to address the real-time control tasks. Despite the kernel of the two modules is the same algorithm, their roles, tasks, and capabilities are quite different. A detailed description of the role of FARM at addressing low-level control tasks is provided, along with tentative operational procedures for training the models embedded into the algorithm by using the collected experimental data.

42 ENGINEERING↗

Finding Horndeski theories with Einstein gravity limits

The Horndeski action is the most general scalar-tensor theory with at most second-order derivatives in the equations of motion, thus evading Ostrogradsky instabilities and making it of interest when modifying gravity at large scales. To pass local tests of gravity, these modifications predominantly rely on nonlinear screening mechanisms that recover Einstein's Theory of General Relativity in regions of high density. We derive a set of conditions on the four free functions of the Horndeski action that examine whether a specific model embedded in the action possesses an Einstein gravity limit or not. For this purpose, we develop a new and surprisingly simple scaling method that identifies dominant terms in the equations of motion by considering formal limits of the couplings that enter through the new terms in the modified action. This enables us to find regimes where nonlinear terms dominate and Einstein's field equations are recovered to leading order. Together with an efficient approximation of the scalar field profile, one can then further evaluate whether these limits can be attributed to a genuine screening effect. For illustration, we apply the analysis to both a cubic galileon and a chameleon model as well as to Brans-Dicke theory. Finally, we emphasise that the scaling method also provides a natural approach for performing post-Newtonian expansions in screened regimes.

79 ASTRONOMY AND ASTROPHYSICS↗