187037 research outputs found
Sort by
Inherent safety evaluation of a long-life fast reactor (SALUS-100) under unprotected loss of heat sink conditions
No full textThe Korea Atomic Energy Research Institute (KAERI) is developing the long-life sodium-cooled fast reactor SALUS-100, designed for continuous 20-year operation without refueling. The concept adapts the proven technology of the Prototype Generation-IV SFR (PGSFR) to a non-light-water small modular reactor platform. To verify that these inherited technologies still provide the intrinsic sodium-cooled fast reactors (SFR) safety features-negative reactivity feedback and passive decay-heat removal via the diverse residual heat-removal system (DRHRS)-an unprotected loss-of-heat-sink (ULOHS) analysis was performed. The transient calculations employed GAMMA+ 2.0, a system code validated against data from JOYO, PFBR, Monju, EBR-II, FFTF and other SFR facilities. A concurrent station blackout was assumed to challenge the passive cooling path, and separate cases examined the loss of one versus two intermediate-loop pumps. During the ULOHS event, a rise in core-inlet temperature triggered dominant negative reactivity through the Doppler effect and radial core expansion, stabilizing the reactor. With a single IHTS pump trip, -0.0207 occurred at 216 s, stabilizing power at 42.6 %. The magnitude of the initial power excursion scaled with the inlettemperature spike, highlighting the influence of rapid temperature changes on early transients. Sensitivity studies also quantified the impact of intermediate heat-transport system (IHTS) pump coast-down characteristics and timing of operator actions.
TFAS: zero-shot NAS for general time-series analysis with time-frequency aware scoring
No full textDesigning effective neural networks from scratch for various time-series analysis tasks, such as activity recognition, fault detection, and traffic forecasting, is time-consuming and heavily relies on human labor. To reduce the reliance on human labor, recent studies adopt neural architecture search (NAS) to design neural networks for time series automatically. Still, existing NAS frameworks for time series only focus on one specific analysis, such as forecasting and classification, with expensive search methods. This paper therefore aims to build a unified zero-shot NAS framework that effectively searches neural architectures for a variety of tasks and time-series data. However, to build a general framework for different tasks, we need a zero-shot proxy that consistently correlates with the downstream performance across different characteristics of time-series datasets. To address these challenges, we propose a zero-shot NAS framework with novel Time-Frequency Aware Scoring for general time-series analysis, named TFAS. To incorporate TFAS into existing foundation time-series models, we adopt time-frequency decomposition methods and introduce the concept of augmented architecture to the foundation models. This augmented architecture enables the zero-shot proxy to be aware of the decomposed time and frequency information, resulting a more accurate estimation of downstream performance for particular datasets. Empirically, we show that the architectures found by TFAS gain improvement of up to 23.6% over state-of-the-art hand-crafted baselines in five mainstream time-series data mining tasks, including short- and long-term forecasting, classification, anomaly detection, and imputation.
Toward Interactive Sound Source Localization: Better Align Sight and Sound!
No full textRecent studies on learning-based sound source localization have primarily focused on localization performance. However, prior work and existing benchmarks often overlook a crucial aspect: cross-modal interaction, which is essential for interactive sound source localization. This interaction is vital for understanding semantically matched or mismatched audio-visual events, such as silent objects or true sound sources among multiple objects. In this work, we comprehensively examine the cross-modal interaction of existing methods, benchmarks, evaluation metrics, and cross-modal understanding tasks. We identify the overlooked points of previous studies and make several contributions to address them. First, we propose a learning framework that incorporates retrieval-based and hand-crafted augmentation techniques, enhancing cross-modal interaction through cross-modal alignment. Second, we introduce new evaluation metrics to accurately and rigorously assess localization methods, focusing on both localization performance and cross-modal interaction. Third, to thoroughly analyze interactive sound source localization, we present a new semi-synthetic benchmark with diverse categorical combinations. Finally, we evaluate both interactive sound source localization and auxiliary cross-modal retrieval tasks, benchmarking competing methods alongside our own. Our new benchmark and evaluation metrics reveal that previous methods struggle with interactive sound source localization tasks, largely due to their limited cross-modal interaction capabilities. Our method, which features enhanced cross-modal alignment, demonstrates superior sound source localization and cross-modal interaction performance. This work provides the most comprehensive analysis of sound source localization to date, with extensive validation of competing methods on both existing and new benchmarks using both new and standard evaluation metrics.
FLOW-AUGMENTATION III: COMPLEXITY DICHOTOMY FOR BOOLEAN CSPS PARAMETERIZED BY THE NUMBER OF UNSATISFIED CONSTRAINTS
No full textWe study the parameterized problem of satisfying ``almost all"" constraints of a given formula T over a fixed, finite Boolean constraint language F, with or without weights. More precisely, for each finite Boolean constraint language F, we consider the following two problems. In MIN SAT(F), the input is a formula T over F and an integer k, and the task is to find an assignment \alpha : V(T)-{0, 1\} that satisfies all but at most k constraints of T, or determine that no such assignment exists. In WEIGHTED MIN SAT(F), the input additionally contains a weight function w:T-Z+ and an integer W, and the task is to find an assignment \alpha such that (1) \alpha satisfies all but at most k constraints of T, and (2) the total weight of the violated constraints is at most W. We give a complete dichotomy for the fixed-parameter tractability of these problems: We show that for every Boolean constraint language F, either WEIGHTED MIN SAT(F) is FPT; or WEIGHTED MIN SAT(F) is W[1[-hard but MIN SAT(F) is FPT; or MIN SAT(F) is W[1[-hard. This generalizes recent work of Kim et al. [in SODA 2021, SIAM, Philadelphia, 2021, pp. 149--168[, which did not consider weighted problems and only considered languages F that cannot express implications (u-v) (as is used to, e.g., model digraph cut problems). Our result generalizes and subsumes multiple previous results, including the FPT algorithms for WEIGHTED ALMOST 2-SAT, weighted and unweighted \ell-CHAIN SAT, and COUPLED MIN-CUT, as well as weighted and directed versions of the latter. The main tool used in our algorithms is the recently developed method of directed flow-augmentation [E. J. Kim et al., in STOC 2022, ACM, 2022, pp. 938--947[.
Techno-economic and environmental impact assessment of feasible alternatives to steam methane reforming for hydrogen production: dry and chemical looping reforming routes
No full textAs hydrogen gains prominence as a key energy carrier, the economic and environmental sustainability of hydrogen production processes has become a critical concern. In this study, dry reforming of methane (DRM) and chemical looping reforming (CLR) were modeled as feasible alternatives to the commercially established steam methane reforming (SMR) process. To evaluate these technologies, a comprehensive analysis was conducted, considering energy consumption, economic feasibility, and environmental impact. CLR demonstrated a 42.6 % reduction in specific energy consumption (SEC) compared to SMR. Regarding cost competitiveness, CLR achieved a 2.73 % reduction in the levelized cost of hydrogen production (LCOH) and provided insights into how the lifespan of metal oxides, which serve as looping solids, affects its economic viability. Although DRM showed a 3.65 % increase in LCOH compared to SMR, an appropriate CO2 feedstock supply could enhance its economic competitiveness. In terms of environmental impact, DRM and CLR reduced net CO2 emissions by 74.8 % and 32.9 % compared to SMR, respectively. These findings highlight that both DRM and CLR present promising alternatives to SMR, offering environmental sustainability in hydrogen production.
Suppression of pressure fluctuation and turbulent friction at the wall in spatially developing turbulent boundary layers via streamwise-traveling waves and wall oscillations
No full textTurbulent boundary layers generate large wall pressure fluctuations and high skin friction drag, which in turn raise aerodynamic and acoustic penalties for aircraft, ships, and wind-energy devices. Active near-wall flow control techniques that target the dominant vortex structures have therefore attracted growing attention as a practical route to simultaneous noise and drag reduction. The impact of flow controls on zero-pressure gradient turbulent boundary layer flow at high Reynolds number was studied using Direct Numerical Simulation to investigate their influence on the altered vortex features and the resulting wall pressure fluctuations. Four distinct boundary layer flow control strategies, blowing (BW), suction (SC), spanwise wall oscillation (SWO), and streamwise-traveling wave with wall oscillation (STWO), were applied to examine their effects on enstrophy, vorticity, wall pressure, and their respective correlations. Turbulent statistics revealed that SC increased skin friction due to enhanced turbulence near the wall, while reductions were seen in BW, SWO, and STWO, with STWO showing the most significant decrease. Analysis of premultiplied energy spectra of the pressure fluctuations in SWO and STWO indicated notable suppression of wall pressure fluctuations and the associated vortex energies generated by the near-wall quasi-streamwise vortices. The correlation between wall pressure fluctuations and streamwise vorticity highlighted the effectiveness of these control methods, particularly with STWO, in redistributing energy and attenuating turbulence structures responsible for wall pressure fluctuations. Flow visualization and enstrophy analysis further confirmed that the control strategies substantially altered vortex structures, demonstrating the potential of SWO and STWO as effective approaches for reducing wall pressure fluctuations and drag mitigation.
Development of High-Thermal-Conductivity, Low-Viscosity Materials for Underfill in Advanced Semiconductor Packaging
No full text
Sensitivity of Pulsatile Parameters of Computed Fractional Flow Reserve using a Reduced-order Model
No full textFractional flow reserve (FFR) is the clinical gold standard for diagnosing coronary artery disease, yet many computational methods rely on steady flow assumptions even though clinical FFR values are measured under pulsatile flow conditions. To address the computational cost and uncertainty associated with pulsatile blood flow simulations, we present a reduced-order model combined with a polynomial chaos expansion (PCE) method. The coronary geometry is represented by thin slabs defined by consecutive centerline points and radius, while a polynomial function approximates pressure differences as a function of flow and its time derivative based on minimal three-dimensional simulations. Diverse pulsatile flow conditions are modeled using lumped parameter models which approximate a wide range of pulsatile flow conditions. Uncertainties in pulsatile parameters, including cardiac output, heart rate, and pulse pressure, are modeled using a third-order Chebyshev PCE to maintain a mean relative error below 1.0%. Validation was conducted using a cylindrical model across stenosis severities from 40% to 90%, as well as a patient-specific model with diverse disease conditions. In both cases, the computed FFR distributions agreed with clinical observations. Sensitivity analysis showed that myocardial compression, distal aortic resistance, and contractility are the primary factors influencing FFR variability, with FFR variation exhibiting a linear correlation with its value. This reduced-order approach enables efficient pulsatile FFR simulations and provides valuable insights into key parameters affecting FFR
