Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “edge inference”

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 109 records · Page 6
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Simulations of tokamak edge plasma turbulent fluctuations based on a minimal 3D model

In this study, a new simulation model for tokamak boundary plasma, SOLT3D, is implemented in the BOUT++ framework. The simulation model includes a set of dynamic equations describing collisional boundary plasma and neutral gas in the tokamak scrape-off layer and divertor region. The model is verified against standard linear plasma instabilities and available nonlinear results. For L-mode like conditions, SOLT3D reproduces characteristics of boundary plasma turbulent fluctuations that are similar to published experimental data, in terms of the amplitude and spatial dependence of the fluctuations. It also reproduces realistic plasma fluxes on material surfaces and expected Bohm-like effective radial transport. Plasma fluctuations inferred from the simulations imply inevitably a significant level of intrinsic 'noise' for axisymmetric tokamak plasma transport modeling, introducing errors on the order of unity. In particular, the toroidally averaged atomic rates below 5–10 eV are strongly modified by turbulent plasma fluctuations, which should significantly affect the standard axisymmetric modeling of the tokamak edge plasma and divertor.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2024

Nanosecond anomaly detection with decision trees and real-time application to exotic Higgs decays

Abstract We present an interpretable implementation of the autoencoding algorithm, used as an anomaly detector, built with a forest of deep decision trees on FPGA, field programmable gate arrays. Scenarios at the Large Hadron Collider at CERN are considered, for which the autoencoder is trained using known physical processes of the Standard Model. The design is then deployed in real-time trigger systems for anomaly detection of unknown physical processes, such as the detection of rare exotic decays of the Higgs boson. The inference is made with a latency value of 30 ns at percent-level resource usage using the Xilinx Virtex UltraScale+ VU9P FPGA. Our method offers anomaly detection at low latency values for edge AI users with resource constraints.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2021

Evaluation of silicon carbide as a divertor armor material in DIII-D H-mode discharges

Silicon carbide (SiC) represents a promising but largely untested plasma-facing material (PFM) for next-step fusion devices. In this work, an analytic mixed-material erosion model is developed by calculating the physical (via SDTrimSP) and chemical (via empirical scalings) sputtering yield from SiC, Si, and C. The Si content in the near-surface SiC layer is predicted to increase during D plasma bombardment due to more efficient physical and chemical sputtering of C relative to Si. Silicon erosion from SiC thereby occurs primarily from sputtering of the enriched Si layer, rather than directly from the SiC itself. SiC coatings on ATJ graphite, manufactured via chemical vapor deposition, were exposed to repeated H-mode plasma discharges in the DIII-D tokamak to test this model. The qualitative trends from analytic modeling are reproduced by the experimental measurements, obtained via spectroscopic inference using the S/XB method. Quantitatively the model slightly under-predicts measured erosion rates, which is attributed to uncertainties in the ion impact angle distribution, as well as the effect of edge-localized modes. After exposure, minimal changes to the macroscopic or microscopic surface morphology of the SiC coatings were observed. Compositional analysis reveals Si enrichment of about 10%, in line with expectations from the erosion model. Extrapolating to a DEMO-type device, an order-of-magnitude decrease in impurity sourcing, and up to a factor of 2 decrease in impurity radiation, is expected with SiC walls, relative to graphite, if low C plasma impurity content can be achieved. Furthermore, these favorable erosion properties motivate further investigations of SiC as a low-Z, non-metallic PFM.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2024

Investigating the effect of B T direction on W source-to-core pathways during the SAS-VW campaign on DIII-D

Experiments using the V-shaped closed slot tungsten (W) coated SAS-VW divertor in DIII-D studied the effects of the BT direction on core contamination of eroded tungsten from a closed slot divertor configuration. Core W content is inferred using soft-X ray tomography (SXR) and vacuum ultraviolet spectroscopy (SPRED), while W divertor erosion is inferred from visible spectroscopy of W emission (400.9 nm) measured by in-slot filterscopes (filtered photo-multipliers). Post-mortem analysis from the campaign discovered tile misalignment leading to suspected pronounced leading-edge erosion in the unfavorable BT direction (ion B ⃗×∇B ⃗ drift away from divertor) likely not captured by diagnostics. However, empirical findings show up to ~2-3x larger core contamination in the favorable BT direction even considering no additional W erosion from leading edges. A “source-to-core efficiency factor” is derived to estimate the effects of leading-edge erosion and compare W contamination for two pairs of H-mode discharges in opposite BT directions. While having differing absolute parameters, similar core impurity density gradients suggest comparable core impurity transport. These results show that favorable BT may have stronger source-to-core pathways for W impurities sourced from the outer divertor region. Possible explanations could include the effects of E ⃗×B ⃗ drifts on W transport in the scrape-off-layer (SOL) as well as previously determined fast SOL inner target directed flows in favorable BT.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2022

Temperature of Suprathermal Electron Strahl at ACE as an Indicator of Solar Coronal Origin Temperature [Slides]

Determine the effectiveness of using electron strahl temperature to infer coronal source temperature. Suprathermal electron strahl temperature derived from measurements at 1 AU does retain information regarding the coronal source temperature from which it originates. This correlation is most strongly measured during the leading edges of coronal hole events but is generally also present in full event measurements. This correlation is also observed in CME events during a coronal hole.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

Boundary, SOL, and Divertor Physics Studies on TCV (Final Report for Research Funded by DoE Award DE-SC0020327)

Research supported under this grant was organized around two diagnostic systems installed on the TCV tokamak located at EPFL in Lausanne, Switzerland: Multi-Spectral Imaging (MSI) and Gas-Puff Imaging (GPI). MSI monitors 2-d time-history images of up to 10 spectral lines, in the visible, using identical views for all wavelengths. The primary application of MSI is to infer 2-d time histories of electron temperature and density in the Scrape-Off-Layer (SOL) of diverted plasmas, utilizing ratios of emission from multiple neutral helium lines. The atomic models are validated against 1-d Thomson scattering measurements. SOL density and temperature maps are analyzed to infer divertor dynamics, including particle and power flows, for a variety of divertor configurations, and specifically as flux expansion along the long-legged outboard divertor is systematically varied. GPI is used to measure details of edge fluctuations in the SOL, particularly as a function of plasma triangularity (δ), including both positive and negative δ. A threshold is found, whereby for δ<-0.3, edge turbulence in the far SOL is completely suppressed. Dynamics of the filamentary turbulence was also analyzed using advance machine learning techniques. Research in support of four PhD theses was carried out within the scope of this Award, including 2 at MIT and 2 at EPFL.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2026

Revealing Local Structures through Machine-Learning-Fused Multimodal Spectroscopy

Atomistic structures of materials offer valuable insights into their functionality. Determining these structures remains a fundamental challenge in materials science, especially for systems with defects. While both experimental and computational methods exist, each has limitations in resolving nanoscale structures. Core-level spectroscopies, such as X-ray absorption (XAS) or electron energy-loss spectroscopies (EELS), have been used to determine the local bonding environment and structure of materials. Recently, machine learning (ML) methods have been applied to extract structural and bonding information from XAS/EELS data. However, frameworks relying solely on a single data stream, defined as characterization data derived from a single element using one technique, are often insufficient because multiple local environments can yield similar spectral features, making it challenging to differentiate between competing structural hypotheses. Here, in this work, we address this challenge by integrating multimodal ab initio simulations, experimental data acquisition, and ML techniques for structure characterization. Our goal is to determine local structures and properties using EELS and XAS data from multiple elements and edges. To showcase our approach, we use various lithium nickel manganese cobalt (NMC) oxide compounds which are used for lithium ion batteries, including those with oxygen vacancies and antisite defects, as the sample material system. We successfully inferred local element content, ranging from lithium to transition metals, with quantitative agreement with experimental data. Beyond local element inference, we find that ML model based on multimodal spectroscopic data is able to determine whether local defects such as oxygen vacancy and antisites are present, a task which is impossible for single mode spectra or other experimental techniques. Furthermore, our framework is able to provide physical interpretability, bridging spectroscopy with the local atomic and electronic structures.

battery↗
DOE OSTI2023

Kinetic simulation of a 50 mTorr capacitively coupled argon discharge over a range of frequencies and comparison to experiments

The effect of driving frequency in the range of 13.56–73 MHz on electron energy distribution and electron heating modes in a 50 mTorr capacitively coupled argon plasma discharge is studied using 1D-3V particle-in-cell simulations. Calculated electron energy probability functions exhibit three distinct “temperatures” for low-, mid-, and high-energy electrons at all the studied driving frequencies. When compared to published experimental data, the calculated probability functions show a reasonable agreement for the energy range resolved in the measurements (about 2–10 eV). Discrepancies due to limitations in experimental energy resolution outside this range lead to differences between computational and experimental values of the electron number density determined from the distribution functions, and the predicted effective electron temperature is within 25% of experimental values. The impedance of the discharge is interpreted in terms of a homogeneous equivalent circuit model, and the driving frequency dependence of the inferred combined sheath thickness is found to obey a known, theoretically derived, power law. The average power transferred from the field to the electrons (electron heating) is computed, and a region of negative heating near the sheath edge, particularly at higher driving frequencies, is identified. Analysis of the electron momentum equation shows that electron inertia, which on temporal averaging would be zero in a linear regime, is responsible for negative values of power deposition near the sheath edge at high driving frequencies due to the highly nonlinear behavior of the discharge.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2022

Accelerating the Inference of the Exa.TrkX Pipeline

Recently, graph neural networks (GNNs) have been successfully used for a variety of particle reconstruction problems in high energy physics, including particle tracking. The Exa.TrkX pipeline based on GNNs demonstrated promising performance in reconstructing particle tracks in dense environments. It includes five discrete steps: data encoding, graph building, edge filtering, GNN, and track labeling. All steps were written in Python and run on both GPUs and CPUs. In this work, we accelerate the Python implementation of the pipeline through customized and commercial GPU-enabled software libraries, and develop a C++ implementation for inferencing the pipeline. The implementation features an improved, CUDA-enabled fixed-radius nearest neighbor search for graph building and a weakly connected component graph algorithm for track labeling. GNNs and other trained deep learning models are converted to ONNX and inferenced via the ONNX Runtime C++ API. The complete C++ implementation of the pipeline allows integration with existing tracking software. We report the memory usage and average event latency tracking performance of our implementation applied to the TrackML benchmark dataset.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2020

Kinetic Modeling and Experimental Studies of Capacitively Coupled Macro- and Microplasmas at Microwave Frequencies

The effect of driving frequency in the range of 13.56 MHz to 73 MHz on electron energy distribution and electron heating modes in a 50 mTorr capacitively coupled argon plasma discharge is studied using 1D-3V particle-in-cell simulations. Calculated electron energy probability functions exhibit three distinct temperatures for low-, mid-, and high-energy electrons at all the studied driving frequencies. When compared to published experimental data, the calculated probability functions show a reasonable agreement for the energy range resolved in the measurements (about 2 eV to 10 eV). Discrepancies due to limitations in experimental energy resolution outside this range lead to differences between computational and experimental values of the electron number density determined from the distribution functions, and the predicted effective electron temperature is within 25 % of experimental values. The impedance of the discharge is interpreted in terms of a homogeneous equivalent circuit model and the driving frequency dependence of the inferred combined sheath thickness is found to obey a known, theoretically-derived, power law. The average power transferred from the field to the electrons (electron heating) is computed, and a region of negative heating near the sheath edge, particularly at higher driving frequencies, is identified. Analysis of the electron momentum equation shows that electron inertia, which on temporal averaging would be zero in a linear regime, is responsible for negative values of power deposition near the sheath edge at high driving frequencies due to the highly nonlinear behavior of the discharge.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2022

Do graph neural networks learn traditional jet substructure?

At the CERN LHC, the task of jet tagging, whose goal is to infer the origin of a jet given a set of final-state particles, is dominated by machine learning methods. Graph neural networks have been used to address this task by treating jets as point clouds with underlying, learnable, edge connections between the particles inside. We explore the decision-making process for one such state-of-the-art network, ParticleNet, by looking for relevant edge connections identified using the layerwise-relevance propagation technique. As the model is trained, we observe changes in the distribution of relevant edges connecting different intermediate clusters of particles, known as subjets. The resulting distribution of subjet connections is different for signal jets originating from top quarks, whose subjets typically correspond to its three decay products, and background jets originating from lighter quarks and gluons. This behavior indicates that the model is using traditional jet substructure observables, such as the number of prongs -- energetic particle clusters -- within a jet, when identifying jets.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2024

Inferring the scrape-off layer heat flux width in a divertor with a low degree of axisymmetry

Plasma facing components (PFCs) in the next generation of tokamak devices will operate in challenging environments, with heat loads predicted to exceed 10 MWm -2 . The magnitude of these heat loads is set by the width of the channel, the ‘scrape-off layer’ (SOL), into which heat is exhausted, and can be characterised by an e-folding length scale for the decay of heat flux across the channel. It is expected this channel will narrow as tokamaks move towards reactor relevant conditions. Understanding the processes involved in setting the SOL heat flux width is imperative to be able to predict the heat loads PFCs must handle in future devices. Measurements of the SOL width are performed on the high-field spherical tokamak, ST40, using a newly commissioned infrared thermography system. With its high on-axis toroidal magnetic field (≥1.5 T) ST40 is uniquely positioned to investigate the influence of toroidal field on the heat flux width in spherical tokamaks, whilst also extending measurements of the SOL width in spherical tokamaks to increased poloidal field (≥0.3 T). Due to the divertor on ST40 having a low degree of axisymmetry, it is necessary for a set of radial measurements of the heat flux to be taken across the divertor, made possible using an automated toolchain that fully incorporates its 3D geometry. These radial profiles are combined with the magnetic geometry of the plasma to infer the width of the SOL, with both Eich and double exponential profiles of heat flux observed. A reduction in the heat flux is observed toroidally across part of the divertor, along with increased heat loads observed locally around the edges of the tiles. Future work in characterising the impact of tile misalignment and uncertainties in the reconstructed divertor magnetic geometry is required in order to further understand the observed heat flux patterns, as are additional investigations into the role potentially being played by an inhomogeneous sheath electric field.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2020

Machine-Learning X-Ray Absorption Spectra to Quantitative Accuracy

Simulations of excited state properties, such as spectral functions, are often computationally expensive and therefore not suitable for high-throughput modeling. As a proof of principle, here we demonstrate that graph-based neural networks can be used to predict the x-ray absorption near-edge structure spectra of molecules to quantitative accuracy. Specifically, the predicted spectra reproduce nearly all prominent peaks, with 90% of the predicted peak locations within 1 eV of the ground truth. Besides its own utility in spectral analysis and structure inference, our method can be combined with structure search algorithms to enable high-throughput spectrum sampling of the vast material configuration space, which opens up new pathways to material design and discovery.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

FPGA Acceleration of GCN in Light of the Symmetry of Graph Adjacency Matrix

Graph Convolutional Neural Networks (GCNs) are widely used to process large-scale graph data. Different from deep neural networks (DNNs), GCNs are sparse, irregular, and unstructured, posing unique challenges to hardware acceleration with regular processing elements (PEs). In particular, the adjacency matrix of a GCN is extremely sparse, leading to frequent but irregular memory access, low spatial/temporal data locality and poor data reuse. Furthermore, a realistic graph usually consists of unstructured data (e.g., unbalanced distributions), creating significantly different processing times and imbalanced workload for each node in GCN acceleration. To overcome these challenges, we propose an end-to-end hardware-software co-design to accelerate GCNs on resource-constrained FPGAs with the features including: (1) A custom dataflow that leverages symmetry along the diagonal of the adjacency matrix to accelerate feature aggregation for undirected graphs. We utilize either the upper or the lower triangular matrix of the adjacency matrix to perform aggregation in GCN to improve data reuse. (2) Unified compute cores for both aggregation and transform phases, with full support to the symmetry-based dataflow. These cores can be dynamically reconfigured to the systolic mode for transformation or as individual accumulators for aggregation in GCN processing. (3) Preprocessing of the graph in software to rearrange the edges and features to match the custom dataflow. This step improves the regularity in memory access and data reuse in the aggregation phase. Moreover, we quantize the GCN precision from FP32 to INT8 to reduce the memory footprint without losing the inference accuracy. We implement our accelerator design in Intel Stratix10 MX FPGA board with HBM2, and demonstrate 1.3x-110.5x improvement in end-to-end GCN latency as compared to the state-of the-art FPGA implementations, on the graph datasets of Cora, Pubmed, Citeseer and Reddit.

Nair, Gopikrishnan R.↗
DOE OSTI2021

Resonant X-ray emission spectroscopy from broadband stochastic pulses at an X-ray free electron laser

Abstract Hard X-ray spectroscopy is an element specific probe of electronic state, but signals are weak and require intense light to study low concentration samples. Free electron laser facilities offer the highest intensity X-rays of any available light source. The light produced at such facilities is stochastic, with spikey, broadband spectra that change drastically from shot to shot. Here, using aqueous ferrocyanide, we show that the resonant X-ray emission (RXES) spectrum can be inferred by correlating for each shot the fluorescence intensity from the sample with spectra of the fluctuating, self-amplified spontaneous emission (SASE) source. We obtain resolved narrow and chemically rich information in core-to-valence transitions of the pre-edge region at the Fe K-edge. Our approach avoids monochromatization, provides higher photon flux to the sample, and allows non-resonant signals like elastic scattering to be simultaneously recorded. The spectra obtained match well with spectra measured using a monochromator. We also show that inaccurate measurements of the stochastic light spectra reduce the measurement efficiency of our approach.

Fuller, Franklin D. (ORCID:0000000237737087)↗
DOE OSTI2026

Characterization of core neutrals using passive measurements of the D α spectrum near the X -point in the DIII-D tokamak

New spectroscopic measurements of deuterium Balmer-α emission are used to infer the spatial distribution of neutral particles near the X-point in a diverted high confinement mode plasma. The charge exchange neutral spectroscopy (CENS) diagnostic—recently installed on the DIII-D tokamak—uses 15 lines of sight extending from the edge of the confined region (ρ ≈ 0.7) to the X-point in lower single-null plasmas. Each CENS line of sight is spectrally resolved around the deuterium Balmer-α wavelength (6561 Å) to measure the Doppler shift, line broadening, and intensity of Dα emission. Thermal line broadening is used to identify emission from neutrals undergoing charge-exchange (CX) with high temperature ions in the confined plasma. This spectral information allows a more accurate determination of the neutral density deeper inside the confined plasma compared to traditional filter-based diagnostics. We present two methods of analyzing CENS measurements: (1) fitting the spectrum from each view-chord separately, and (2) a tomographic method for inverting the neutral density over a 2D region of space using all CENS views collectively. The neutral density profile is found to decay exponentially in the radial direction with two scale-lengths. In the pedestal region the neutral density decays at a rate approximately equal to the local mean-free-path for CX collisions, $L_{n_\mathrm{D0}}\approx \lambda_\text{CX}$. Further inside the plasma the neutral density decays at a rate equal to the mean-free-path for ionization, $L_{n_\mathrm{D0}}\approx \lambda_\text{inz.}$. The separatrix $n_\mathrm{D0}$ value is found to be approximately $2\times10^{15}$ m−3. Based on results from the 2D inversion, the density of neutrals is found to increase along the separatrix approaching the X-point by a factor of $\approx 5\times$ over the region covered by the CENS diagnostic.

X-point neutrals↗
DOE OSTI2022

Edge termination in vertical GaN diodes: Electric field distribution probed by second harmonic generation

We characterized the electric field distribution of GaN-on-GaN p–n diodes with partially compensated ion-implanted edge termination (ET) using an electric field induced second harmonic generation technique (EFISHG). The distributed electric field from the anode to the outer edge of the ET demonstrates the effectiveness of the ET structure. However, EFISHG also shows that its effectiveness is strongly dependent on the acceptor charge distribution in the ET's partially compensated layer (PC). A generally lower amount of acceptor charge can be inferred from the measured electric field distribution resulting from excessive ion implantation energy or dose during ET fabrication and causing lower than optimal breakdown voltage. Localized field crowding can be observed when the remaining acceptors uncompensated by the implant in the PC layer are nonuniformly distributed around the periphery of the devices. Important information can be obtained from these direct electric field measurements and used for optimizing the device design and fabrication process.

Cao, Yuke↗
DOE OSTI2021

3D radiation, density, and MHD structures following neon shattered pellet injection into stable DIII-D Super H-mode discharges

Six nominally repeat neon shattered pellet injection (SPI) shutdowns of stable DIII-D Super H-modes are studied to understand the 3D properties of the radiation and impurity transport. The radiation efficiency and radiation peaking determine whether first wall melting is expected following disruption mitigation in ITER. Previous studies make use of axisymmetric approximations to infer radiation efficiencies, but validating the high efficiency required by ITER necessitates improved accuracy, and this work contributes by exploring the 3D radiation and density structures that will inform forward modeling. When the neon shatter plume produced by the SPI reaches the plasma edge, m/n = 3/1 and 2/1 island O-points are observed to align with the injection trajectory in five out of six cases, suggesting that the injected material seeds the island O-points. Field aligned neon structures emitting Ne-I line radiation drift at 1 km/s in the ion diamagnetic drift direction during the pre-TQ, tracking the motion of the m/n = 2/1 island O-point. Neon fragments penetrate to the q = 2 surface by the time of the TQ. Techniques to constrain the 3D emissivity are explored, and one method constrains a 3D flux tube that is consistent with the radiation data, and when mapped to the interferometers, intersects the lasers that measure the highest density. The resulting structure derived from the radiation measurements exists near the 2/1 island X-point. In five repeatable discharges, the peak of the radiation in the toroidal direction exists in a 120° toroidal sector where the injection occurs, in contrast with the outlier discharge where the toroidal peak exists in the complementary 240° toroidal sector far from the injector, and where a 50% lower density rise is observed. In conclusion, the n = 1 phase behavior is markedly different in the outlier discharge, suggesting a possible dependence of the radiation structure and the assimilation efficiency on MHD.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗