Search NASA⌕ Search

SEARCH · Search NASA

Results for “Computing methodologies”

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 991 records · Page 55

V&V Framework

A Verification and Validation (V&V) framework is presented for the development and execution of coordinated modeling and experimental program s to assess the predictive capability of computational models of complex systems through focused, well structured, and formal processes. The elements of the framework are based on established V&V methodology developed by various organizations including the Department of Energy, National Aeronautics and Space Administration, the American Institute of Aeronautics and Astronautics, and the American Society of Mechanical Engineers. Four main topics are addressed: 1) Program planning based on expert elicitation of the modeling physics requirements, 2) experimental design for model assessment, 3) uncertainty quantification for experimental observations and computational model simulations, and 4) assessment of the model predictive capability. The audience for this document includes program planners, modelers, experimentalist, V &V specialist, and customers of the modeling results.

97 MATHEMATICS AND COMPUTING↗

Used Nuclear Fuel Loading and Structural Performance Under Normal Conditions of Transport - Modeling, Simulation and Experimental Integration RD&D Plan

Under current U.S. Nuclear Regulatory Commission regulation, it is not sufficient for used nuclear fuel (UNF) to simply maintain its integrity during the storage period, it must maintain its integrity in such a way that it can withstand the physical forces of handling and transportation associated with restaging the fuel and moving it to treatment or recycling facilities, or a geologic repository. Hence it is necessary to understand the performance characteristics of aged UNF cladding and ancillary components under loadings stemming from transport initiatives. Researchers would like to demonstrate that enough information, including experimental support and modeling and simulation capabilities, exists to establish a preliminary determination of UNF structural performance under normal conditions of transport (NCT). This research, development and demonstration (RD&D) plan describes a methodology, including development and use of analytical models, to evaluate loading and associated mechanical responses of UNF rods and key structural components. This methodology will be used to provide a preliminary assessment of the performance characteristics of UNF cladding and ancillary components under rail-related NCT loading. The methodology couples modeling and simulation and experimental efforts currently under way within the Used Fuel Disposition Campaign (UFDC). The methodology will involve limited uncertainty quantification in the form of sensitivity evaluations focused around available fuel and ancillary fuel structure properties exclusively. The work includes collecting information via literature review, soliciting input/guidance from subject matter experts, performing computational analyses, planning experimental measurement and possible execution (depending on timing), and preparing a variety of supporting documents that will feed into and provide the basis for future initiatives. The methodology demonstration will focus on structural performance evaluation of Westinghouse WE 17×17 pressurized water reactor fuel assemblies with a discharge burnup range of 30-58 GWd/MTU (assembly average), loaded in a representative high-capacity (≥32 fuel rod assemblies) transportation package. Evaluations will be performed for representative normal conditions of rail transport involving a rail conveyance capable of meeting the Association of American Railroads (AAR) S-2043 specification. UNF modeling is anticipated to be defined to the pellet-cladding level and take in to account influences associated with spacer grids, intermediate fluid mixers, and control components. The influence of common degradation issues such as ductile-to-brittle-transition will also be accounted for. All model development and analysis will be performed with commercially available software packages exclusively. Inputs and analyses will be completely documented, all supporting information will be traceable, and bases will be defendable so as to be most useful to the U.S. Department of Energy community and mission. The expected completion date is the end of fiscal year (FY) 2013.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Parallel replica dynamics method for bistable stochastic reaction networks: Simulation and sensitivity analysis

Stochastic reaction networks that exhibit bistable behavior are common in systems biology, materials science, and catalysis. Sampling of stationary distributions is crucial for understanding and characterizing the long-time dynamics of bistable stochastic dynamical systems. However, simulations are often hindered by the insufficient sampling of rare transitions between the two metastable regions. In this paper, we apply the parallel replica method for a continuous time Markov chain in order to improve sampling of the stationary distribution in bistable stochastic reaction networks. The proposed method uses parallel computing to accelerate the sampling of rare transitions. Furthermore, it can be combined with the path-space information bounds for parametric sensitivity analysis. With the proposed methodology, we study three bistable biological networks: the Schlögl model, the genetic switch network, and the enzymatic futile cycle network. We demonstrate the algorithmic speedup achieved in these numerical benchmarks. More significant acceleration is expected when multi-core or graphics processing unit computer architectures and programming tools such as CUDA are employed

97 MATHEMATICS AND COMPUTING↗

Environmental and Energy Research at the Texas Center for Advanced Scientific Computing and Modeling (CASCaM)

The funding enabled the creation of the (University of North) Texas Center for Advanced Scientific Computing and Modeling (CASCaM). Due to its success, CASCaM has continued to grow beyond the initial team focused on the biological and environmental research discussed herein, with investments by the University of North Texas in new faculty and researchers who contribute to CASCaM research. CASCaM is now a sustainable effort, with researchers across multiple fields including chemistry, physics, biology, mechanical engineering, materials science, and other fields, with multiple researchers engaged in a broad variety of research, including other DOE projects. Over 20 PI’s and ~100 researchers are engaged in CASCaM research. The overall goal of the research effort was to provide greater understanding about carbon capture, storage, and utilization. As part of this effort, the development of computational methods and identification of novel strategies to support the CO2 and CO studies were necessary. Three primary themes were investigated: carbon dioxide and CO catalysis; protein- and amino acid-based strategies towards carbon dioxide capture, storage, and utilization; and, finally, the methodology improvements for modeling carbon dioxide chemistry

54 ENVIRONMENTAL SCIENCES↗

ASSESSMENT OF MULTIPHASE CFD WITH ZERO CLOSURE MODEL FOR BOILING WATER REACTOR FUEL ASSEMBLIES

Extension of existing boiling closures to the high void fraction conditions of BWRs is one of challenges for M-CFD applications. This paper presents the assessment of an Eulerian M-CFD model with Zero closure model that has been specifically designed for two-phase flow and heat transfer applications in BWR fuel assemblies. The baseline Zero closure aims at including the simplest and most robust representation of the key physical mechanisms, rather than aiming at absolute accuracy of the predictions. The capabilities of M-CFD with Zero were evaluated against the international OECD/NRC BFBT benchmark data through the comparative analyses of local void fraction, sub-channel void fraction error and exit quality. INL high performance computing system has been employed to demonstrate the scalability of the M-CFD application to nuclear reactor fuel assemblies. The M-CFD simulation results show a good agreement with the experimental data for both quantitative and qualitative results, with the highest uncertainty being observed in the prediction of the local void fraction profile at low void fraction condition. This paper presents a recent view of the best practices for M-CFD of BWR fuel assemblies to provide a useful reference to identify and overcome common challenges in the applications of M-CFD tools. We provide recommendations for geometric modeling, mesh generation and solver configurations to optimize the methodology application.

97 MATHEMATICS AND COMPUTING↗

Silicon Qubits

There are two good reasons to attempt to build quantum bits (qubits) out of silicon. The first is the obvious foundation of classical microelectronics. Although silicon quantum computers would operate in a fundamentally different way from classical computers$-$for example, at cryogenic temperatures$-$still the level of development in material quality, crystal growth, and fabrication methodologies for silicon is unrivaled by any other material in the world. Leveraging even a small fraction of the worldwide investment in silicon for qubit development could potentially put silicon-based qubits far ahead of other solid-state alternatives. The second, less obvious reason for choosing silicon is the remarkably clean magnetic environment witnessed by spins in highly purified and isotopically enriched silicon material. Fortuitously, 95.3% of the naturally occurring isotopes of Si nuclei ( 28 Si and 30 Si) are spin-0. These nuclei therefore have a “closed shell” of nuclear moments, providing no external magnetic field whatsoever. Add to this the possibility of intrinsic silicon with part-per-billion chemical quality and the system is remarkably close to “vacuum” with respect to magnetic noise properties.

36 MATERIALS SCIENCE↗

Metastable Materials Discovery in the Age of Large-Scale Computation

Computational materials discovery has been successful in predicting novel, technologically relevant materials. However, it has remained focused almost exclusively on finding ground-state structures. Now that the lower-hanging fruit has been found in many fields of application, materials exploration is moving toward metastable materials: higher energy phases that are stable at practical time scales. Because of the challenges associated with predicting which phases are realistic, this class of materials has remained relatively unexplored, despite numerous examples of metastable structures with unmatched properties (e.g., diamond). This article highlights recent advances in developing computational and theoretical methods for predicting useful and realizable metastable materials. Topics discussed cover (1) the latest strategies for identifying potential metastable phases, (2) methodologies for assessing which phases can be realized experimentally, and (3) current approaches to estimate the lifetime of metastable materials.

amorphous materials↗

Impact of the differential fluence distribution of brachytherapy sources on the spectroscopic dose-rate constant

Purpose: To investigate why dose-rate constants for {sup 125}I and {sup 103}Pd seeds computed using the spectroscopic technique, Λ{sub spec}, differ from those computed with standard Monte Carlo (MC) techniques. A potential cause of these discrepancies is the spectroscopic technique’s use of approximations of the true fluence distribution leaving the source, φ{sub full}. In particular, the fluence distribution used in the spectroscopic technique, φ{sub spec}, approximates the spatial, angular, and energy distributions of φ{sub full}. This work quantified the extent to which each of these approximations affects the accuracy of Λ{sub spec}. Additionally, this study investigated how the simplified water-only model used in the spectroscopic technique impacts the accuracy of Λ{sub spec}. Methods: Dose-rate constants as described in the AAPM TG-43U1 report, Λ{sub full}, were computed with MC simulations using the full source geometry for each of 14 different {sup 125}I and 6 different {sup 103}Pd source models. In addition, the spectrum emitted along the perpendicular bisector of each source was simulated in vacuum using the full source model and used to compute Λ{sub spec}. Λ{sub spec} was compared to Λ{sub full} to verify the discrepancy reported by Rodriguez and Rogers. Using MC simulations, a phase space of the fluence leaving the encapsulation of each full source model was created. The spatial and angular distributions of φ{sub full} were extracted from the phase spaces and were qualitatively compared to those used by φ{sub spec}. Additionally, each phase space was modified to reflect one of the approximated distributions (spatial, angular, or energy) used by φ{sub spec}. The dose-rate constant resulting from using approximated distribution i, Λ{sub approx,i}, was computed using the modified phase space and compared to Λ{sub full}. For each source, this process was repeated for each approximation in order to determine which approximations used in the spectroscopic technique affect the accuracy of Λ{sub spec}. Results: For all sources studied, the angular and spatial distributions of φ{sub full} were more complex than the distributions used in φ{sub spec}. Differences between Λ{sub spec} and Λ{sub full} ranged from −0.6% to +6.4%, confirming the discrepancies found by Rodriguez and Rogers. The largest contribution to the discrepancy was the assumption of isotropic emission in φ{sub spec}, which caused differences in Λ of up to +5.3% relative to Λ{sub full}. Use of the approximated spatial and energy distributions caused smaller average discrepancies in Λ of −0.4% and +0.1%, respectively. The water-only model introduced an average discrepancy in Λ of −0.4%. Conclusions: The approximations used in φ{sub spec} caused discrepancies between Λ{sub approx,i} and Λ{sub full} of up to 7.8%. With the exception of the energy distribution, the approximations used in φ{sub spec} contributed to this discrepancy for all source models studied. To improve the accuracy of Λ{sub spec}, the spatial and angular distributions of φ{sub full} could be measured, with the measurements replacing the approximated distributions. The methodology used in this work could be used to determine the resolution that such measurements would require by computing the dose-rate constants from phase spaces modified to reflect φ{sub full} binned at different spatial and angular resolutions.

07 ISOTOPE AND RADIATION SOURCES↗

Sintering-resistant Single-Site Nickel Catalyst Supported by Metal-Organic Framework

Developing supported single-site catalysts is an important goal in heterogeneous catalysis, since the well-defined active sites afford opportunities for detailed mechanistic studies, thereby facilitating the design of improved catalysts. We present herein a method for installing Ni ions uniformly and precisely on the node of a Zr-based MOF, NU-1000, in high density and large quantity (denoted as Ni-AIM) using atomic layer deposition (ALD) in a metal–organic framework (MOF) (AIM). Ni-AIM is demonstrated to be an efficient gas-phase hydrogenation catalyst upon activation. The structure of the active sites in Ni-AIM is proposed, revealing its single-site nature. More importantly, due to the organic linker used to construct the MOF support, the Ni ions stay isolated throughout the hydrogenation catalysis, in accord with its long-term stability. A quantum chemical characterization of the catalyst and the catalytic process complements the experimental results. With validation of computational modeling protocols, we further targeted ethylene oligomerization catalysis by Ni-AIM guided by theoretical prediction. Given the generality of the AIM methodology, this emerging class of materials should prove ripe for the discovery of new catalysts for the transformation of volatile substrates.

Li, Zhanyong↗

Nuclear data uncertainty propagation and modeling uncertainty impact evaluation in neutronics core simulation

Uncertainty analysis is a critical requirement in reactor simulation as it is used to quantify the reliability of best-estimate calculation. A comprehensive uncertainty analysis should characterize all sources of uncertainties in a computationally-feasible and scientifically-defendable manner. Here we employ a well-established reduced order modeling (ROM) based uncertainty quantification methodology to propagate uncertainties throughout neutronic calculations. ROM relies on recent advances in randomized data mining techniques applied to large data streams. In our proposed implementation, the nuclear data uncertainties are first propagated from multi-group level through lattice physics calculation to generate few-group parameter uncertainties, described using a vector of mean values and a covariance matrix. Employing an ROM-based compression of the covariance matrix, the few-group uncertainties are then propagated through downstream core simulation in a computationally efficient manner. This straightforward approach, albeit efficient as compared to brute force forward and/or adjoint-based methods, often employs a number of assumptions that have been unquestioned in the literature of neutronic uncertainty analysis. This manuscript argues that these assumptions could introduce another source of uncertainty referred to as modeling uncertainties, whose magnitude needs to be quantified in tandem with nuclear data uncertainties. Thus, our primary goal is to explore the interactions between these two uncertainty sources in order to assess whether modeling uncertainties have an impact on parameter uncertainties. To explore this endeavor, the impact of a number of modeling assumptions on core attributes uncertainties is quantified. The study employs a CANDU reactor model, with Serpent and NEWT as lattice physics solvers and NESTLE-C as core simulator. The modeling assumptions investigated include those related with the uncertainty propagation method employed, e.g., deterministic vs. stochastic, the few-group energy structure employed to represent the cross-sections, the resonance treatment in lattice physics calculation, the reference values for the cross-section, and the number of samples employed to render ROM compression. Results indicate that some of the modeling assumptions could have a non-negligible impact on the core responses propagated uncertainties, highlighting the need for a more comprehensive approach to combine parameter and modeling uncertainties.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Towards a reference numerical scheme using MCNPX for PWR control rod tip fluence estimations

Recent occurrences of cracks and fissures on the cladding tubes of PWR control rod (CR) fingers employed in the Swiss reactors prompted the need to develop more reliable analytical methods for CR tip fluence estimations. To partly address this need, a deterministic methodology based on SIMULATE-3/CASMO-4 was in recent years developed at PSI. Although this methodology has already been applied for independent support to licensing issues related to CR lifetime, two main questions are currently being the center of attention for further enhancements. First, the methodology relies on several assumptions that have so far not been verified. Secondly, an assessment of the achieved accuracy has not been addressed. In an attempt to answer both these open questions, it was considered appropriate to develop an alternative computational scheme based on the stochastic MCNPX code with the objective to provide reference numerical solutions. This paper presents the first steps undertaken in that direction. To start, a methodology for a volumetric neutron source transfer to full core MCNPX models with detailed CR as well as axial reflector representations is established. On this basis, the assumptions of the deterministic methodology are studied for selected CR configurations for two Beginning-of-Life cores by comparing the spatial neutron flux distributions obtained with the two approaches for the entire spectrum. Finally, for the high-energy range (E> 1 MeV) and for a few CRs, the new MCNPX scheme is applied to estimate the accumulated fluence over one real operated cycle and the results are compared with the deterministic approach. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

The Dynamics of Ions on Phased Radio-frequency Carpets in High Pressure Gases and Application for Barium Tagging in Xenon Gas Time Projection Chambers

Radio-frequency (RF) carpets with ultra-fine pitches are examined for ion transport in gases at atmospheric pressures and above. We develop new analytic and computational methods for modeling RF ion transport at densities where dynamics are strongly influenced by buffer gas collisions. An analytic description of levitating and sweeping forces from phased arrays is obtained, then thermodynamic and kinetic principles are used to calculate ion loss rates in the presence of collisions. This methodology is validated against detailed microscopic SIMION simulations. We then explore a parameter space of special interest for neutrinoless double beta decay experiments: transport of barium ions in xenon at pressures from 1 to 10 bar. Our computations account for molecular ion formation and pressure dependent mobility as well as finite temperature effects. We discuss the challenges associated with achieving suitable operating conditions, which lie beyond the capabilities of existing devices, using presently available or near-future manufacturing techniques.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Report on the Workshop on Accelerated Nuclear Energy Materials Development

This document reports on the Office of Nuclear Energy’s (NE’s) Workshop on Accelerated Nuclear Energy Materials Development held May 11, 2010, in Washington, DC. The purpose of the workshop was twofold: (1) to provide feedback on an initiative to use uncertainty quantification (UQ) to integrate theory, simulation, and modeling with accelerated experimentation to predict the behavior of materials and fuels in an irradiation environment and thereby accelerate the lengthy materials design and qualification process; and (2) to provide feedback on and refinement to five topical areas to develop predictive models for fuels and cladding and new radiation-tolerant materials. The goal of the workshop was to gather technical feedback with respect to the Office of Nuclear Energy’s research and development while also identifying and highlighting crosscutting capability and applicability of the initiative to other federal offices, including the Department of Energy’s (DOE’s) National Nuclear Security Administration (NNSA), Nuclear Regulatory Commission (NRC), DOE Office of Basic Energy Sciences (BES), DOE Office of Fusion Energy Sciences (FES), and Naval Reactors. The goals of the initiative are twofold: (1) develop time- and length-scale transcending models that predict material properties using UQ to effectively integrate theory, simulation, and modeling with accelerated experiments; and (2) design and develop new radiation-tolerant materials using the knowledge gained and methodologies created to shorten the development and qualification time and reduce cost. The initiative is crosscutting and has synergy with industry and other federal offices including Naval Reactors, NRC, FES, BES, and the Office of Advanced Scientific Computing Research (ASCR). It is distinguished by its use of uncertainty quantification to effectively integrate theory, simulation, and modeling with high-dose experimental capabilities. The initiative aims to bring the methodology that is being successfully applied in NNSA’s Stockpile Stewardship Program to accelerate the development of advanced nuclear energy materials. Industry, Naval Reactors, and NRC expressed support for this type of initiative and a general interest in collaboration in the area of light water reactor materials research.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Market Driven Residential Energy Codes: Comparing Performance in a Changing Technological Environment

The research project is undertaken to better understand the changing relationship between the two basic methods of building energy code compliance – prescriptive and performance – and how those methods relate to each other with respect to advancements in building energy computer simulation standards and capabilities. The International Energy Efficiency Code (IECC) is a model code adopted by many jurisdictions across the United States. Historically, the prescriptive compliance methodology has been preferred in most jurisdictions. The prescriptive methodology requires meeting or exceeding specific efficiency minimums for each envelope component. This tends to be a simple method to teach and verify. A more involved prescriptive alternative called the Total UA alternative is sometimes used. This method requires some multiplication, summing, and comparison to compute, so it is done with a fairly simple computer program. However, advances in computer and building energy simulation technology have resulted in increased use of more detailed performance compliance methods. The performance compliance method establishes the annual energy cost threshold via hourly simulation models. The compliance threshold is determined with a comparison building model simulation with geometry similar to the proposed home and with energy feature parameters and efficiencies as specified in the IECC. This project examines relationships between the two methods of building energy code compliance, including: • Overall annual energy use based on utility bill analysis by compliance method • Code official work processes with respect to compliance methods • Gaps and issues associated with building code compliance methods • Simulated energy use difference between compliance methods • Code compliance cost as a function of compliance method • Code compliance labeling effectiveness for high performance residences • Getting to net zero energy use and net zero greenhouse gas emissions through high performance code alternatives • Electronic code permitting and compliance alternatives

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

A Methodology to Evaluate the Grid Reliability Impact of Oscillations Induced by Large Loads

The rapid growth of hyperscale AI data centers is bringing renewed attention to the reliability risk that sustained forced oscillations pose to bulk power systems, with cyclic computational workloads emerging as a new forcing source. Unlike the broadband, stochastic disturbances from traditional industrial loads such as arc furnaces, AI training and inference facilities can inject large active power swings concentrated at specific frequencies over extended durations - characteristics that existing grid planning practices do not account for. While the North American Electric Reliability Corporation (NERC) has recognized this gap and called for system-level studies of large load interconnections, no standardized methodology exists to screen, simulate, and quantify these risks at the planning stage. This report presents the Risk Assessment Tool for Large Load-induced Events (RATLLE), a Python-based, publicly available script suite developed at the Pacific Northwest National Laboratory to evaluate bulk power system reliability risks from data center-induced oscillations. RATLLE implements a three-module workflow: a screening module that identifies vulnerable interconnection locations and excitable system modes; a simulation module that models cyclic data center load behavior using a commercial positive sequence simulation platform; and an analysis module that computes risk metrics and generates interactive visualization dashboards. The risk metrics, formulated around simulation observables, map oscillation impacts to a three-stage severity scale spanning latent equipment fatigue through imminent cascading failure. The methodology is demonstrated on two Western Electricity Coordinating Council (WECC) system models: a publicly available 240-bus reduced representation and a detailed 2031 Heavy Winter planning case. Case studies illustrate that even modest 50 MW forced oscillations at resonant frequencies can produce wide-area power swings, N-1 security constraint violations, and cascading generator trips through protection actions - outcomes that would not occur under normal operating conditions without oscillations present. The results underscore the need for standardized oscillation impact assessment in large load interconnection studies and provide a reproducible, extensible framework for utilities to adopt or customize within their existing planning workflows.

Biswas, Shuchismita↗

An effective thermal-hydraulics methodology for prismatic core HTGR and VHTR

Optimizing the performance and design of a prismatic core HTGR or VHTR requires a full core thermal-hydraulics analysis. Owing to the complexity and massive core structure, such analysis requires extensive and massively parallelized computation capabilities and a relatively long time (weeks to months) to complete. These demanding requirements are not due to the 3-D simulation of heat conduction in the annular core of the reactor, but rather the 3-D computational fluid dynamics (CFD) simulation of the helium gas flow in the 10-m long cooling channels in the 102 hexagonal fuel elements and the axial graphite reflector blocks in the core. This paper applies and examines the effectiveness of using a 1-D simulation of the helium flow in the core coolant channels, coupled to a 3-D simulation of the heat conduction in the graphite and fuel compacts, to perform thermal-hydraulics analysis of a hexagonal fuel element and of a 1/6 full core. This methodology employs typical cosine and constant axial power profiles and an applicable convective heat transfer correlation for the helium flow in the coolant channels. The correlation has recently been validated for a 10 m tall, single channel fuel module and shown to significantly reduce the computation time and memory requirements without compromising the accuracy of the calculations. The fidelity and accuracy of the present results for a hexagonal fuel element are verified by comparing them to those of a full 3-D numerical analysis. In addition to the temperature fields, results compare the computation time and number of numerical grid elements for implementing the two numerical simulation methods. The results of the thermal-hydraulics analysis of a 1/6 full core with the simplified methodology are also presented. All performed analysis accounts for the temperature dependent properties of helium, graphite in the reactor core and reflector blocks and the TRISO particle fuel compacts. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Developing tools and process controls to manufacture energy-efficient powders for additive manufacturing feedstocks: Computational analysis of metal powder manufacturing via machining

Traditionally, metal powders have been produced through methods such as grinding, atomization, and electrolysis. In contrast to these techniques, Metal Powder Works, Inc. has pioneered a methodology based on metal cutting. This innovative approach utilizes a vibrating cutting tool to machine metal particles, in the form of chips, from a workpiece. This technique allows for control of powder particle size, morphology, and avoids any thermally induced material changes. This collaboration aims to elucidate metal cutting characteristics and assess performance on tough materials like Inconel alloys. Computational models, using FEA and SPH techniques, will be developed initially, focusing on aluminum alloy (Al 7075-T6) for studying mesh sensitivity, cutting forces, and chip morphology.

99 GENERAL AND MISCELLANEOUS↗

SGD-Net: Efficient Model-Based Deep Learning with Theoretical Guarantees

Deep unfolding networks have recently gained popularity for solving imaging inverse problems. However, the computational and memory complexity of data-consistency layers within traditional deep unfolding networks scales with the number of measurements, limiting their applicability to large-scale imaging inverse problems. We propose SGD-Net as a new methodology for improving the efficiency of deep unfolding through stochastic approximations of the data-consistency layers. Our theoretical analysis shows that SGD-Net can be trained to approximate batch deep unfolding networks to an arbitrary precision. Our simulations on intensity diffraction tomography and sparse-view computed tomography show that SGD-Net can match the performance of the traditional batch network at a fraction of training and testing complexity.Deep unfolding networks have recently gained popularity for solving imaging inverse problems. However, the computational and memory complexity of data-consistency layers within traditional deep unfolding networks scales with the number of measurements, limiting their applicability to large-scale imaging inverse problems. We propose SGD-Net as a new methodology for improving the efficiency of deep unfolding through stochastic approximations of the data-consistency layers. Our theoretical analysis shows that SGD-Net can be trained to approximate batch deep unfolding networks to an arbitrary precision. Our simulations on intensity diffraction tomography and sparse-view computed tomography show that SGD-Net can match the performance of the traditional batch network at a fraction of training and testing complexity.

97 MATHEMATICS AND COMPUTING↗