187037 research outputs found
Sort by
Damage modeling of high-crimp carbon/phenolic woven composites incorporating weft yarn straightening-induced matrix cracking
No full textCarbon/phenolic (CP) composites, particularly rayon-based CP woven composites, are widely utilized in aircraft and rocket engine nozzles due to their excellent thermal resistance. However, existing progressive damage models cannot adequately capture the unique failure mechanisms in high-crimp rayon-based fabrics. This study introduces a novel weft yarn straightening matrix cracking failure mode, motivated by experimental observations, to capture matrix cracking and modulus transformation induced by high crimp (crimp angle 40 degrees, crimp ratio 9.1 %). Unlike existing models, this mode distinctly explains the coupled mechanism of weft-direction modulus transformation and through-thickness degradation. The proposed model incorporates plasticity theory and continuum damage mechanics to capture the gradual load transfer mechanism during fiber straightening. Material properties were determined through comprehensive mechanical testing, and three-point bending validation was performed on both on-axis and 45 degrees off-axis specimens. For on-axis tests, predicted values were 616.2 N and 1,429mJ versus experimental 608.3 N and 1,510mJ. For 45 degrees off-axis tests, predicted values were 299.1 N and 1,365mJ versus experimental 300.7 N and 1,311mJ. The proposed model exhibits accuracy and superior performance compared to existing approaches, particularly for composites with high fiber crimp.
Hierarchical Planning for Vehicle Routing and Scheduling in Marsupial Robotic Systems
No full textThis letter presents a hierarchical planning approach to the vehicle routing and scheduling problem (VRSP) for marsupial robotic systems, a specialized class of heterogeneous robotic systems in which one type of mobile robot is capable of carrying another. While traditional VRSPs have been widely studied, the marsupial variant (MVRSP) has received relatively little attention. To address the NP-hard nature of MVRSP, this work introduces a hierarchical planning structure that decomposes the problem into two subproblems with reduced complexity: a high-level routing problem, formulated as a mixed-integer linear program (MILP), and a low-level scheduling problem, modeled in the Planning Domain Definition Language (PDDL). These subproblem solutions are integrated to generate complete mission plans. The proposed approach is validated through qualitative plan visualizations and quantitative Monte Carlo simulations in an autonomous subsea mapping scenario, where an unmanned surface vehicle carries multiple underwater vehicles. Results show that the hierarchical planner significantly improves both planning efficiency and solution quality compared to baseline methods.
On the real zeros of depth 1 quasimodular forms
No full textWe discuss the critical points of modular forms, or more generally the zeros of quasimodular forms of depth 1 for PSL2(Z). In particular, we consider the derivatives of the unique weight k modular forms fk with the maximal number of consecutive zero Fourier coefficients following the constant 1. Our main results state that (1) every zero of a depth 1 quasimodular form near the derivative of the Eisenstein series in the standard fundamental domain lies on the geodesic segment {z is an element of H : R(z) = 1/2}, and (2) more than quarter of zeros of fk in the standard fundamental domain lie on the geodesic segment {z is an element of H:R(z) = 1/2} for large enough k with k equivalent to 0 (mod 12). (c) 2025 Elsevier Inc. All rights are reserved, including those for text and data mining, AI training, and similar technologies.
Adaptive integration of textual context and visual embeddings for underrepresented vision classification
No full textThe advancement of deep learning has significantly improved image classification performance; however, handling long-tail distributions remains challenging due to the limited data available for rare classes. Existing approaches predominantly focus on visual features, often neglecting the valuable contextual information provided by textual data, which can be especially beneficial for classes with sparse visual examples. In this work, we introduce a novel method addressing this limitation by integrating textual data generated by advanced language models with visual inputs through our newly proposed Adaptive Integration Block for Vision-Text Synergy (AIB-VTS). Specifically designed for Vision Transformer architectures, AIB-VTS adaptively balances visual and textual information during inference, effectively utilizing textual descriptions generated from large language models. Extensive experiments on benchmark datasets demonstrate substantial performance improvements across all class groups, particularly in underrepresented (tail) classes. These results confirm the effectiveness of our approach in leveraging textual context to mitigate data scarcity issues and enhance model robustness.
Cross-ratio and vehicle dynamics-based speed estimation for traffic accident analysis
No full textIn traffic accident analysis, vehicle speed estimation is crucial for determining accident causation and establishing legal liability. Although cross-ratio-based methods are widely employed in traffic accident video analysis, they remain primarily limited to average speed calculations on straight road sections. This study proposes a video-based speed estimation technique applicable to curved roads by integrating cross-ratio geometric principles with vehicle dynamics. The proposed method estimates continuous speed variations from video data through automatic selection of reliable frame combinations. Accuracy is further enhanced by incorporating vehicle dynamics principles specific to curved driving scenarios. The method was validated through comparative analyses with existing approaches, employing PC-Crash simulations and real accident cases. Experimental results demonstrate that the proposed method improves speed estimation accuracy on curved roads and during acceleration or deceleration. The method also proves effective for the temporal analysis of event data recorder (EDR) data in multiple-collision accidents. This video-based approach is expected to enhance the reliability and objectivity of traffic surveillance video analysis, particularly for application as legal evidence.
Joint graphical lasso with regularized aggregation
No full textWe present methods for estimating multiple precision matrices for high-dimensional time series within the framework of Gaussian graphical models, with a specific focus on analyzing functional magnetic resonance imaging (fMRI) data collected from multiple subjects. Our goal is to estimate both individual brain networks and a collective structure representing a group of subjects. To achieve this, we propose a method that utilizes group Graphical Lasso and regularized aggregation to simultaneously estimate individual and group precision matrices, assigning varying weights to each individual based on their outlier status within the group. We investigate the convergence rates of precision matrix estimators under various norms and expectations, assessing their performance with sub-Gaussian and heavy-tailed data. The effectiveness of our methods is demonstrated through simulations and real fMRI data analysis.
Finely tunable thermal expansion of NiTi by stress-induced martensitic transformation and thermomechanical training
No full textTailoring the thermal expansion of martensitic materials by crystallographic texture and anisotropic variation of lattice parameters is a promising route to the flexible design of thermally stable systems. NiTi alloys are prototype materials in this respect, exhibiting shape-memory and superelastic properties owing to their thermoelastic martensitic transformations. Here, we propose a method to realize finely tunable coefficients of thermal expansion (CTE) for the NiTi alloy based on a special combination of mechanical and thermal training. We achieve a near-zero in-plane CTE smaller than that of the FeNi-based Invar alloy. Atomistic simulations and theoretical calculations guide the method design and clarify the underlying mechanisms of the relationship between the processing conditions, the microstructural evolution, and the thermal expansion behavior. The directions for further, finer adjustments of the CTE without constraints on the shape of the materials are indicated.
Investigation of a data-driven hybrid machine learning model for nuclear proliferation risk prediction for early warning
No full textEarly detection of nuclear proliferation risk is inherently challenging due to the rarity and secrecy of the underlying escalation events. This study presents an investigation of a data-driven early-warning model based on historical data by transforming discrete proliferation stages into a continuous 0-to-1 risk score. The model uses political-economic and HS-code trade variables for 148 countries from 1939 to 2012. The model architecture (i) applies supervised learning regression (LightGBM regressors) in rolling windows to track annual stage scores based on Bleek and Narang taxonomies, (ii) detects residual anomalies using unsupervised learning via Isolation Forests, and (iii) fuses these signals in a meta-classifier to generate interpretable yearly alarm probabilities. The final model achieves an event-F1 score of 0.65 and ROC-AUC of 0.99, with a mean warning lead time of +1.14 years against the historical proliferation events. Quantifying the contributions of inputs to predictions using SHAP analysis reveals a post-1995 shift in proliferation drivers toward uranium centrifuge-related trade patterns. This underscores the need to examine the growing role of global supply chains. The system model based on an explainable, tree-based ensemble machine learning technique provides a transparent, lead-time-positive alternative to "black-box" deep learning and presents the possibility of developing a methodology for integrating multi-modal AI and dynamic data streams into an early warning tool to support nonproliferation intelligence.
Spatial control of silane layer formation on boron nanoparticles via diverse methods for high-energetic nanofuel applications
No full textThis study presents a spatial control strategy of silane layer thickness using injection and dropwise methods to enhance the dispersion stability and combustion efficiency of energetic boron nanoparticles (BNs) in hydro-carbon fuels. Dropwise addition enables uniform silane coating through controlled hydrolysis and condensation,while rapid injection results in deposits of low-density siloxane on BNs. The dropwise silane-coated reduced BNs (DrBNOs) exhibit improved dispersion stability, low sedimentation, and enhanced compatibility with hydro-carbon fuels. Additionally, the long-chain silane coating effectively passivates nanoparticle surface, thereby increasing oxidation resistance and enabling faster ignition, complete combustion, and optimized energy release. Bomb calorimetry and droplet combustion analysis confirmed that DrBNOs deliver high calorific values with minimal residual soot. These findings highlight the importance of silane layer thickness in maximizing com-bustion performance, with the spatial control strategy offering a promising approach for developing advanced energetic nanoadditives for scramjet engine fuels.
Salience theory and stock returns: The role of reference-dependent preferences
No full textSalience theory suggests that investors find assets with salient upsides appealing and those with salient downsides unappealing. Our study shows that this effect is reference-dependent, and particularly strong among stocks with prior capital losses. Using a reference-dependent preference framework, we find that investors with prior losses tend to prefer high-salience stocks, reflecting a risk-loving behavior. This salience effect is pronounced in loss regions, muted or reversed in gain regions, and more pronounced among stocks with low institutional ownership and high arbitrage risk, which aligns with individual investor behavior. The effect is amplified during periods of positive-sentiment, high market volatility, and high uncertainty. Our findings are validated through robustness tests and subsample analyses.