Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Iterative limiters”

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 397 records · Page 22
All sourcesNASA NTRSDOE OSTI
NASA NTRS2016

NDARC NASA Design and Analysis of Rotorcraft - Input, Appendix 4

The NDARC code performs design and analysis tasks. The design task involves sizing the rotorcraft to satisfy specified design conditions and missions. The analysis tasks can include off-design mission performance analysis, flight performance calculation for point operating conditions, and generation of subsystem or component performance maps. The principal tasks (sizing, mission analysis, flight performance analysis) are shown in the figure as boxes with heavy borders. Heavy arrows show control of subordinate tasks. The aircraft description consists of all the information, input and derived, that denes the aircraft. The aircraft consists of a set of components, including fuselage, rotors, wings, tails, and propulsion. This information can be the result of the sizing task; can come entirely from input, for a fixed model; or can come from the sizing task in a previous case or previous job. The aircraft description information is available to all tasks and all solutions. The sizing task determines the dimensions, power, and weight of a rotorcraft that can perform a specified set of design conditions and missions. The aircraft size is characterized by parameters such as design gross weight, weight empty, rotor radius, and engine power available. The relations between dimensions, power, and weight generally require an iterative solution. From the design flight conditions and missions, the task can determine the total engine power or the rotor radius (or both power and radius can be fixed), as well as the design gross weight, maximum takeoff weight, drive system torque limit, and fuel tank capacity. For each propulsion group, the engine power or the rotor radius can be sized. Missions are defined for the sizing task, and for the mission performance analysis. A mission consists of a number of mission segments, for which time, distance, and fuel burn are evaluated. For the sizing task, certain missions are designated to be used for design gross weight calculations; for transmission sizing; and for fuel tank sizing. The mission parameters include mission takeoff gross weight and useful load. For specified takeoff fuel weight with adjustable segments, the mission time or distance is adjusted so the fuel required for the mission equals the takeoff fuel weight. The mission iteration is on fuel weight or energy. Flight conditions are specified for the sizing task, and for the flight performance analysis. For the sizing task, certain flight conditions are designated to be used for design gross weight calculations; for transmission sizing; for maximum takeoff weight calculations; and for anti-torque or auxiliary thrust rotor sizing. The flight condition parameters include gross weight and useful load. For flight conditions and mission takeoff, the gross weight can be maximized, such that the power required equals the power available. A flight state is defined for each mission segment and each flight condition. The aircraft performance can be analyzed for the specified state, or a maximum effort performance can be identified. The maximum effort is specified in terms of a quantity such as best endurance or best range, and a variable such as speed, rate of climb, or altitude.

Rotorcraft Input↗
NASA NTRS2022

NDARC - NASA Design and Analysis of Rotorcraft: Input and Data Structures - Vol 2

The NDARC code performs design and analysis tasks. The design task involves sizing the rotorcraft to satisfy specified design conditions and missions. The analysis tasks can include off-design mission performance analysis, flight performance calculation for point operating conditions, and generation of subsystem or component performance maps. Figure 1-1 illustrates the tasks. The principal tasks (sizing, mission analysis, flight performance analysis) are shown in the figure as boxes with heavy borders. Heavy arrows show control of subordinate tasks. The aircraft description (figure 1-1) consists of all the information, input and derived, that defines the aircraft. The aircraft consists of a set of components, including fuselage, rotors, wings, tails, and propulsion. This information can be the result of the sizing task; can come entirely from input, for a fixed model; or can come from the sizing task in a previous case or previous job. The aircraft description information is available to all tasks and all solutions (indicated by light arrows). The sizing task determines the dimensions, power, and weight of a rotorcraft that can perform a specified set of design conditions and missions. The aircraft size is characterized by parameters such as design gross weight, weight empty, rotor radius, and engine power available. The relations between dimensions, power, and weight generally require an iterative solution. From the design flight conditions and missions, the task can determine the total engine power or the rotor radius (or both power and radius can be fixed), as well as the design gross weight, maximum takeoff weight, drive system torque limit, and fuel tank capacity. For each propulsion group, the engine power or the rotor radius can be sized. Missions are defined for the sizing task, and for the mission performance analysis. A mission consists of a number of mission segments, for which time, distance, and fuel burn are evaluated. For the sizing task, certain missions are designated to be used for design gross weight calculations; for transmission sizing; and for fuel tank sizing. The mission parameters include mission takeoff gross weight and useful load. For specified takeoff fuel weight with adjustable segments, the mission time or distance is adjusted so the fuel required for the mission (burned plus reserve) equals the takeoff fuel weight. The mission iteration is on fuel weight or energy. Flight conditions are specified for the sizing task, and for the flight performance analysis. For the sizing task, certain flight conditions are designated to be used for design gross weight calculations; for transmission sizing; for maximum takeoff weight calculations; and for antitorque or auxiliary thrust rotor sizing. The flight condition parameters include gross weight and useful load. For flight conditions and mission takeoff, the gross weight can be maximized, such that the power required equals the power available. A flight state is defined for each mission segment and each flight condition. The aircraft performance can be analyzed for the specified state, or a maximum effort performance can be identified. The maximum effort is specified in terms of a quantity such as best endurance or best range, and a variable such as speed, rate of climb, or altitude. The aircraft must be trimmed, by solving for the controls and motion that produce equilibrium in the specified flight state. Different trim solution definitions are required for various flight states. Evaluating the rotor hub forces may require solution of the blade flap equations of motion.

NDARC↗
NASA NTRS2022

NDARC - NASA Design and Analysis of Rotorcraft: Input - Vol 3

The NDARC code performs design and analysis tasks. The design task involves sizing the rotorcraft to satisfy specified design conditions and missions. The analysis tasks can include off-design mission performance analysis, flight performance calculation for point operating conditions, and generation of subsystem or component performance maps. Figure 1-1 illustrates the tasks. The principal tasks (sizing, mission analysis, flight performance analysis) are shown in the figure as boxes with heavy borders. Heavy arrows show control of subordinate tasks. The aircraft description (figure 1-1) consists of all the information, input and derived, that defines the aircraft. The aircraft consists of a set of components, including fuselage, rotors, wings, tails, and propulsion. This information can be the result of the sizing task; can come entirely from input, for a fixed model; or can come from the sizing task in a previous case or previous job. The aircraft description information is available to all tasks and all solutions (indicated by light arrows). The sizing task determines the dimensions, power, and weight of a rotorcraft that can perform a specified set of design conditions and missions. The aircraft size is characterized by parameters such as design gross weight, weight empty, rotor radius, and engine power available. The relations between dimensions, power, and weight generally require an iterative solution. From the design flight conditions and missions, the task can determine the total engine power or the rotor radius (or both power and radius can be fixed), as well as the design gross weight, maximum takeoff weight, drive system torque limit, and fuel tank capacity. For each propulsion group, the engine power or the rotor radius can be sized. Missions are defined for the sizing task, and for the mission performance analysis. A mission consists of a number of mission segments, for which time, distance, and fuel burn are evaluated. For the sizing task, certain missions are designated to be used for design gross weight calculations; for transmission sizing; and for fuel tank sizing. The mission parameters include mission takeoff gross weight and useful load. For specified takeoff fuel weight with adjustable segments, the mission time or distance is adjusted so the fuel required for the mission (burned plus reserve) equals the takeoff fuel weight. The mission iteration is on fuel weight or energy. Flight conditions are specified for the sizing task, and for the flight performance analysis. For the sizing task, certain flight conditions are designated to be used for design gross weight calculations; for transmission sizing; for maximum takeoff weight calculations; and for antitorque or auxiliary thrust rotor sizing. The flight condition parameters include gross weight and useful load. For flight conditions and mission takeoff, the gross weight can be maximized, such that the power required equals the power available. A flight state is defined for each mission segment and each flight condition. The aircraft performance can be analyzed for the specified state, or a maximum effort performance can be identified. The maximum effort is specified in terms of a quantity such as best endurance or best range, and a variable such as speed, rate of climb, or altitude. The aircraft must be trimmed, by solving for the controls and motion that produce equilibrium in the specified flight state. Different trim solution definitions are required for various flight states. Evaluating the rotor hub forces may require solution of the blade flap equations of motion.

NDARC↗
NASA NTRS2023

NDARC - NASA Design and Analysis of Rotorcraft: Input - Appendix 3

The NDARC code performs design and analysis tasks. The design task involves sizing the rotorcraft to satisfy specified design conditions and missions. The analysis tasks can include off-design mission performance analysis, flight performance calculation for point operating conditions, and generation of subsystem or component performance maps. Figure 1-1 illustrates the tasks. The principal tasks (sizing, mission analysis, flight performance analysis) are shown in the figure as boxes with heavy borders. Heavy arrows show control of subordinate tasks. The aircraft description (figure 1-1) consists of all the information, input and derived, that defines the aircraft. The aircraft consists of a set of components, including fuselage, rotors, wings, tails, and propulsion. This information can be the result of the sizing task; can come entirely from input, for a fixed model; or can come from the sizing task in a previous case or previous job. The aircraft description information is available to all tasks and all solutions (indicated by light arrows). The sizing task determines the dimensions, power, and weight of a rotorcraft that can perform a specified set of design conditions and missions. The aircraft size is characterized by parameters such as design gross weight, weight empty, rotor radius, and engine power available. The relations between dimensions, power, and weight generally require an iterative solution. From the design flight conditions and missions, the task can determine the total engine power or the rotor radius (or both power and radius can be fixed), as well as the design gross weight, maximum takeoff weight, drive system torque limit, and fuel tank capacity. For each propulsion group, the engine power or the rotor radius can be sized. Missions are defined for the sizing task, and for the mission performance analysis. A mission consists of a number of mission segments, for which time, distance, and fuel burn are evaluated. For the sizing task, certain missions are designated to be used for design gross weight calculations; for transmission sizing; and for fuel tank sizing. The mission parameters include mission takeoff gross weight and useful load. For specified takeoff fuel weight with adjustable segments, the mission time or distance is adjusted so the fuel required for the mission (burned plus reserve) equals the takeoff fuel weight. The mission iteration is on fuel weight or energy. Flight conditions are specified for the sizing task, and for the flight performance analysis. For the sizing task, certain flight conditions are designated to be used for design gross weight calculations; for transmission sizing; for maximum takeoff weight calculations; and for antitorque or auxiliary thrust rotor sizing. The flight condition parameters include gross weight and useful load. For flight conditions and mission takeoff, the gross weight can be maximized, such that the power required equals the power available. A flight state is defined for each mission segment and each flight condition. The aircraft performance can be analyzed for the specified state, or a maximum effort performance can be identified. The maximum effort is specified in terms of a quantity such as best endurance or best range, and a variable such as speed, rate of climb, or altitude. The aircraft must be trimmed, by solving for the controls and motion that produce equilibrium in the specified flight state. Different trim solution definitions are required for various flight states. Evaluating the rotor hub forces may require solution of the blade flap equations of motion.

NDARC↗
DOE OSTI2022

Deconvolving the roles of E × B shear and pedestal structure in the energy confinement quality of super H-mode experiements

Analysis of “super H-mode” experiments on DIII-D has put forward that high plasma toroidal rotation, not high pedestal, plays the essential role in achieving energy confinement quality H 98y2 >>1. Recently, super H-mode experiments with variable input torque have confirmed that high rotation shear discharges have very high levels of H 98y2 (>1.5), independent of the pedestal height, and that high pedestal discharges with low rotation shear have levels of H 98y2 only slightly above 1 (≤1.2). Although some increase in stored energy with higher pedestal occurs, the energy confinement quality mainly depends on the toroidal rotation shear, which varies according to different levels of injected neutral beam torque per particle. Here, quasi-linear gyrofluid modeling achieves a good match of the experiment when including the E×B shear; without including plasma rotation, the modeling predicts a confinement quality consistent with the empirical observation of H 98y2 ~1.2 at low rotation. Nonlinear gyrokinetic transport modeling shows that the effect of E×B turbulence stabilization is far larger than other mechanisms, such as the so-called hot-ion stabilization (T i /T e ) effect. Consistent with these experimental and modeling results are previous simulations of the ITER Baseline Scenario using a super H-mode pedestal solution, which showed the potential to exceed the Q=10 target if the pedestal density could be increased above the Greenwald limit. A close look at these simulations reveals that the predicted energy confinement quality is below 1 even at the highest pedestal pressure. The improvement in Q at higher pedestal density is due to the improved fusion power generation at the higher core density associated with higher pedestal density, not to an improved energy confinement quality.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2023

An empirical method for geometric calibration of a photon counting detector-based cone beam CT system

BACKGROUND: Geometric calibration is essential in developing a reliable computed tomography (CT) system. It involves estimating the geometry under which the angular projections are acquired. Geometric calibration of cone beam CTs employing small area detectors, such as currently available photon counting detectors (PCDs), is challenging when using traditional-based methods due to detectors’ limited areas. OBJECTIVE: This study presented an empirical method for the geometric calibration of small area PCD-based cone beam CT systems. METHODS: Unlike the traditional methods, we developed an iterative optimization procedure to determine geometric parameters using the reconstructed images of small metal ball bearings (BBs) embedded in a custom-built phantom. An objective function incorporating the sphericities and symmetries of the embedded BBs was defined to assess performance of the reconstruction algorithm with the given initial estimated set of geometric parameters. The optimal parameter values were those which minimized the objective function. The TIGRE toolbox was employed for fast tomographic reconstruction. To evaluate the proposed method, computer simulations were carried out using various numbers of spheres placed in various locations. Furthermore, efficacy of the method was experimentally assessed using a custom-made benchtop PCD-based cone beam CT. RESULTS: Computer simulations validated the accuracy and reproducibility of the proposed method. The precise estimation of the geometric parameters of the benchtop revealed high-quality imaging in CT reconstruction of a breast phantom. Within the phantom, the cylindrical holes, fibers, and speck groups were imaged in high fidelity. The CNR analysis further revealed the quantitative improvements of the reconstruction performed with the estimated parameters using the proposed method. CONCLUSION: Apart from the computational cost, we concluded that the method was easy to implement and robust.

Instruments & Instrumentation↗
DOE OSTI2022

Reassessing energy deposition for the ITER 5 MA vertical displacement event with an improved DINA model

The beryllium (Be) main chamber wall interaction during a 5 MA/1.8 T upward, unmitigated VDE scenario, first analysed in [J. Coburn et al., Phys. Scr. T171 (2020) 014076] for ITER, has been re-evaluated using the latest energy deposition analysis software. Updates to the DINA disruption model are summarized, including an improved numerical convergence for the 0D power balance, limitations on the safety factor within the plasma core, and the choice to maintain a constant plasma +halo poloidal cross-section. Such updates result in a broad halo region and higher radiated power fractions compared to previous models. The new scenario lasts for ~75 ms and deposits ~29 MJ of energy, with the radial distribution of parallel heat flux q‖(r)resembling an exponential falloff with an effective λ_q=75-198 mm. A maximum halo width w_h of 0.52 m at the outboard midplane is observed. SMITER field line tracing and energy deposition simulations calculate a q_{⊥,max} of ~83 MW/m^2 on the upper first wall panels (FWP). Heat transfer calculations with the MEMOS-U code show that the FWP surface temperature reaches ~1000 K, well below the Be melt threshold. Variations of this 5 MA scenario with Be im-purity densities from 0 to 3∙10^19 m^-3 also remain below the melt threshold despite differences in energy deposition and duration. These results are in contrast to the early study which predicted melt damage to the first wall [J. Coburn et al., Phys. Scr. T171 (2020) 014076], and emphasize the importance of accurate models for the halo width w_h and the heat flux distribution q‖(r)within that halo width. The 2020 halo model in DINA has been compared with halo current experiments on COMPASS, JET, and Alcator C-Mod, and the preliminary results build confidence in the broad halo width predictions. Results for the 5 MA VDE are compared with those for a 15 MA equivalent, generated using the new DINA model. At the higher current, significant melting of the upper FWP is to be expected.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
NASA NTRS1979

An approach to eliminate roundoff errors in digital filters

'Second-order quantizers' are introduced which can be used for implementing recursive digital filters with practically no roundoff errors or limit-cycle oscillations. Based on the idea of changing the transfer function used to compute roundoff errors, these quantizers save the low-order bits to correct the product at future iterations. For several pole locations, this can be done with a minimal amount of extra hardware.

Abu-El-haija, A. I.↗
NASA NTRS1982

Mathematical enhancement of data from scientific measuring instruments

The accuracy of any physical measurement is limited by the instruments performing it. The proposed activities of this grant are related to the study of and application of mathematical techniques of deconvolution. Two techniques are being investigated: an iterative method and a function continuation Fourier method. This final status report describes the work performed during the period July 1 to December 31, 1982.

Ioup, J. W.↗
NASA NTRSDate not supplied

High-Fidelity CFD Verification Workshop 2024 Summary: Spalart-Allmaras QCR2000-R Turbulence Model

This paper summarizes solutions submitted for the Reynolds-averaged Navier-Stokes (RANS) test suite of the High-Fidelity CFD Verification Workshop. The goal of the workshop is to establish standards for verification of computational fluid dynamics (CFD) approaches to simulations of steady and unsteady turbulent flows. The RANS verification studies focus on a one-equation Spalart-Allmaras model with quadratic constitutive relation and rotation correction, SA-neg-QCR2000-R. The verification test cases are a two-dimensional subsonic flow around a Joukowski airfoil, a three-dimensional subsonic flow around an extruded NACA 0012 wing in a tunnel, and a subsonic flow around a wing-body configuration developed for verification of solvers participating in the 5 𝑡 ℎ High-Lift Prediction Workshop. The turbulencemodel formulation, geometry, flow conditions, grids, and reference solutions are described in detail. Solutions for the test cases are computed by seven established CFD solvers on adaptedand fixed-grid families using different discretization approaches. While some noticeable differences between solutions remain, the results achieved by contributing solvers show that different solutions computed for the same RANS model on different grid families can converge to a common limit with grid refinement. The apparent requirements for grid convergence are a well designed family of grids that provide sufficient resolution in important areas and a strong solver capable of deep iterative convergence on each grid. For each test case in the study, the variation between aerodynamic forces computed by different solvers on the finest grids of different families is less than 2%.

Boris Diskin↗
DOE OSTI2026

Double-real corrections to color singlet decay in a parton-shower inspired scheme

We introduce a local infrared subtraction method for next-to-next-to-leading order QCD calculations in color singlet decays, with counterterms based on scalar radiators and pure splitting functions. Overlapping singularities in the multipole radiation pattern are disentangled by partial fractioning, and the kinematics mapping corresponds to iterated next-to-leading order kinematics. We verify that the double-real remainder to $e^+e^-\to\;q\bar{q}$ is rendered finite in the single and double unresolved limits and investigate the numerical convergence of the Monte-Carlo integral. We compute the phase-space integrals of the scalar counterterms in the back-to-back configuration, both analytically and with the help of numerical techniques based on sector decomposition.

Campbell, John M. [Fermilab]↗
DOE OSTI2026

Chemical and Radiological Compatibility Testing of 3D Printed Materials

To dramatically increase the adaptability, performance, and safety of processes in support of the Defense Waste Processing Facility (DWPF), Savannah River National Lab (SRNL) plans to perform chemical and radiological compatibility testing on a wide variety of 3D printed materials of interest. The 3D printing process provides numerous strategic operational benefits such as rapid prototyping of complex designs and geometry specific to the needs of the nuclear waste disposition process, as well as on-demand rapid prototyping and iteration with materials that aren’t as accessible through traditional manufacturing methods. Reaction chemistry in simulated waste batches can be matched closely to its radioactive counterpart, but glass reactor vessels have limitations. Vessel geometry can play a big factor in mixing transport limitations, process chemistry, and degradation reaction kinetics. In addition, additive manufacturing allows for much more detailed vessel design than traditional alternatives. Waste processing techniques in DWPF also encounter extreme chemical environments including high pH, strong acids, abrasive slurries, and significant irradiation. To meet these challenges, a matrix of various polymer, ceramic, and metal additive manufacturing materials have been exposed to a suite of chemical environments of interest as well as radioactive dose (such as gamma radiation from 60 Co) to properly test their durability under these conditions. Mass change has been monitored over a period of up to a week in these conditions, as well as added characterization for surface modification through Scanning Electron Microscopy/Electron Dispersive X-ray analysis (SEM/EDX). Further chemical characterization has been monitored through Fourier-Transform InfraRed Spectroscopy (FTIR), with planned investigation via thermal and tensile strength degradation. While the direct product of this research is identification of material(s) that can withstand specific hazardous environments encountered by the mercury water wash tank in DWPF process simulation experiments, the reference base of materials will be used for many other nuclear processes in the pursuit of rapidly developed, cost-efficient, and highly specific devices for environmental remediation and much more.

Wilson, Nathan W. [Savannah River National Laborat↗
DOE OSTI2022

Joint iterative reconstruction and 3D rigid alignment for X-ray tomography

X-ray tomography is widely used for three-dimensional structure determination in many areas of science, from the millimeter to the nanometer scale. The resolution and quality of the 3D reconstruction is limited by the availability of alignment parameters that correct for the mechanical shifts of the sample or sample stage for the images that constitute a scan. In this paper we describe an algorithm for marker-free, fully automated and accurately aligned and reconstructed X-ray tomography data. Our approach solves the tomographic reconstruction jointly with projection data alignment based on a rigid-body deformation model. We demonstrate the robustness of our method on both synthetic phantom and experimental data and show that our method is highly efficient in recovering relatively large alignment errors without prior knowledge of a low resolution approximation of the 3D structure or a reasonable estimate of alignment parameters.

36 MATERIALS SCIENCE↗
DOE OSTI2021

Characteristics of Ultralow Aspect Ratio Toroidal Plasmas and Non-Solenoidal Startup and Edge Stability Studies at Near-Unity Aspect Ratio (Final Report)

This grant supported a long-term research and development program to explore the physics and technology of spherical tokamak (ST) plasma behavior at near-unity aspect ratio ( A ). The PEGASUS Toroidal Experiment is a university-scale ( R ) ~ 0.40 m, α ~ .35 m, I p ≤ 0.3 MA, B T ≤ 0.15 T) experiment that accesses A → 1, to provide large values of the normalized plasma current, I N = I p /α B T , or equivalently the toroidal field utilization factor I p / I TF . The research activities focused on: 1) developing the technique of Local Helicity Injection (LHI) as a viable means of initiating and growing an ST plasma without the use of a central solenoid; 2) exploring the equilibrium and stability properties of spherical tokamak plasmas at extremely low aspect ratio; and 3) obtaining new insights in the properties of the enhanced confinement H-mode regime at A ~ 1. The initial (Phase I) facility employed a unique high-stress central solenoid Ohmic for current drive and heating with simple magnet current waveforms. A subsequent major upgrade (Phase II) offered more flexible control of the plasma evolution along with a powerful LHI system for plasma initiation, heating and sustainment. The LHI technique consists of injection of high-density current streams into vacuum toroidal and vertical magnetic fields. These streams become unstable and support magnetic reconnection to allow the system to relax into a tokamak-like magnetic configuration. This program focused on: 1) development of electron current sources that provide sufficiently high injection currents in the challenging edge plasma region; 2) validation of theoretical limits to achievable I p arising from magnetic helicity and energy conservation plus absolute limits arising from relaxation to a minimal energy states; 3) tests of a power balance model to describe the evolution of the plasma current I p (t); and 4) study of magnetic fluctuations and their role in the I p evolution. This culminated in the production of toroidal currents of I p ≥ 0.2 MA with only ~8 kA of injected current. Operation with Ohmic induction alone in Phase I created plasmas with soft limits on the achievable I p and field utilization ( I p / I TF ~ 1) due to the onset of large low-order (2/1 or 3/2) internal resonant tearing modes. With the expanded capabilities in Phase II operation, methods employed to mitigate those modes and achieve I p / I TF > 1 were: 1) I TF ramp-downs; 2) non-solenoidal startup via LHI; and 3) extremely low I TF Ohmic operations assisted by LHI current injector pre-ionization. Extensive use of LHI-driven startup and drive for non-solenoidal operation produced favorable current profiles to stably operate tokamak plasmas with unprecedentedly high I N = I p / aB T ~ 15 and I p / I TF > 2. With strong reconnection-driven anomalous ion heating, this in turn provided access to tokamak plasmas with B t up to 100%, a world record 2 times higher than that previously achieved in advanced or spherical tokamaks. This high B t regime had a large fraction (~50%) of the plasma volume in a region of an absolute minimum- B configuration. The MHD stability limit in this regime was the ideal external kink mode without a stabilizing wall. These results provided the first demonstration of access to tokamak plasmas with very high B t , high i N , high elongation κ, and low internal inductance ℓ i , and as such served as a proof-ofprinciple for access to this long-sought ST regime. The naturally low B T needed for stability at A ≤ 1.2 reduces the L-H mode power threshold, P LH , so that H-mode plasmas were achieved with Ohmic heating alone. The power threshold for the transition to the H-mode regime was found to be 10× the accepted values from the ITER international scaling, confirming initial trends indicated by experiments on MAST and NSTX, and suggesting that the physics behind this transition is not yet fully understood. The density dependence of the transition and the observed insensitivity to whether the plasma is bounded by a limiter or magnetic divertor are consistent with the recent FM 3 model. Multiple unstable toroidal magnetic modes were measured during ELM (Edge Localized Mode) crashes, with generally lower toroidal mode numbers than seen at higher A , which was attributed to the larger peeling mode instability drive at low A . First-ever edge current profile measurements through an ELM crash showed current-carrying filaments were generated, as seen in ELM simulations and in electromagnetic blob transport theory.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2025

Persistent Sampling: Enhancing the Efficiency of Sequential Monte Carlo

Sequential Monte Carlo (SMC) samplers are powerful tools for Bayesian inference but suffer from high computational costs due to their reliance on large particle ensembles for accurate estimates. We introduce persistent sampling (PS), an extension of SMC that systematically retains and reuses particles from all prior iterations to construct a growing, weighted ensemble. By leveraging multiple importance sampling and resampling from a mixture of historical distributions, PS mitigates the need for excessively large particle counts, directly addressing key limitations of SMC such as particle impoverishment and mode collapse. Crucially, PS achieves this without additional likelihood evaluations-weights for persistent particles are computed using cached likelihood values. This framework not only yields more accurate posterior approximations but also produces marginal likelihood estimates with significantly lower variance, enhancing reliability in model comparison. Furthermore, the persistent ensemble enables efficient adaptation of transition kernels by leveraging a larger, decorrelated particle pool. Experiments on high-dimensional Gaussian mixtures, hierarchical models, and non-convex targets demonstrate that PS consistently outperforms standard SMC and related variants, including recycled and waste-free SMC, achieving substantial reductions in mean squared error for posterior expectations and evidence estimates, all at reduced computational cost. PS thus establishes itself as a robust, scalable, and efficient alternative for complex Bayesian inference tasks.

Karamanis, Minas↗
DOE OSTI2024

Interpretable, extensible linear and symbolic regression models for charge density prediction using a hierarchy of many-body correlation descriptors

Here, density functional theory (DFT) is routinely used to make electronic structure predictions for high-throughput screening of materials and molecules for technologically relevant areas, like the identification of better catalysts, electronic materials, and drug discovery. However, the DFT formalism is limited by (a) its poor (quadratic-to-quartic) scaling, and (b) the need to perform repeated eigenvalue computations of the electronic Hamiltonian as part of its self-consistent field (SCF) iteration procedure to obtain the converged ground state electron density, ρ (r). Approaches that directly predict ρ (r) of a structure with high accuracy can accelerate conventional SCF calculations and can also be used in linearly scaling methods such as orbital-free DFT. To this end, we present a procedure to predict the ground state electron density of molecular and periodic three-dimensional systems directly from the atomic structure with a particular emphasis on physical interpretability. In our framework, ρ (r) is modeled using many-body correlation descriptors that accurately capture the effects of local atomic arrangements in the neighborhood of a grid point. Our use of a linear regression scheme to fit to charge density data enables transparent analysis of the relative contributions of various types of local atomic correlations. By systematically including increasingly complex correlations, our model is shown to accurately predict ρ (r) for a variety of chemically and electronically diverse systems — amorphous Ge, Al(001) slab, crystalline Ga 2 O 3 , molecular benzene, and polyethylene. We then demonstrate a symbolic regression-based protocol to construct easily computable, interpretable features from lower-order correlations that significantly improves our electron density predictions with effectively no increase in the computational cost.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2021

An adaptive Hessian approximated stochastic gradient MCMC method

Bayesian approaches have been successfully integrated into training deep neural networks. One popular family is stochastic gradient Markov chain Monte Carlo methods (SG-MCMC), which have gained increasing interest due to their ability to handle large datasets and the potential to avoid overfitting. Although standard SG-MCMC methods have shown great performance in a variety of problems, they may be inefficient when the random variables in the target posterior densities have scale differences or are highly correlated. Here, we present an adaptive Hessian approximated stochastic gradient MCMC method to incorporate local geometric information while sampling from the posterior. The idea is to apply stochastic approximation (SA) to sequentially update a preconditioning matrix at each iteration. The preconditioner possesses second-order information and can guide the random walk of a sampler efficiently. Instead of computing and saving the full Hessian of the log posterior, we use limited memory of the samples and their stochastic gradients to approximate the inverse Hessian-vector multiplication in the updating formula. Moreover, by smoothly optimizing the preconditioning matrix via SA, our proposed algorithm can asymptotically converge to the target distribution with a controllable bias under mild conditions. To reduce the training and testing computational burden, we adopt a magnitude-based weight pruning method to enforce the sparsity of the network. Our method is user-friendly and demonstrates better learning results compared to standard SG-MCMC updating rules. The approximation of inverse Hessian alleviates storage and computational complexities for large dimensional models. Numerical experiments are performed on several problems, including sampling from 2D correlated distribution, synthetic regression problems, and learning the numerical solutions of heterogeneous elliptic PDE. The numerical results demonstrate great improvement in both the convergence rate and accuracy.

97 MATHEMATICS AND COMPUTING↗