Search Results (6,309)

Convolutional neural networks (CNN) have transformed visual recognition, yet robust geometric reasoning, reliable out-of-distribution generalization, and recognition from limited data remain substantially unsolved. CNNs draw their architectural inspiration from the mammalian visual cortex, but the translation from biology to engineering was selective and, in places, imprecise, and those imprecisions have consequences that are well documented. This paper examines where the biological fidelity holds and where it gives way, grounding the analysis in formal results that predate deep learning and in recent empirical findings on CNN failure modes. We identify three diagnosable architectural limitations. First, CNNs conflate visual modalities that the biological system separates structurally at the lateral geniculate nucleus, feeding raw RGB pixels into a single undifferentiated filter bank and entangling orientation, color, and texture signals from the first layer onward. Second, CNNs repeat a spatial subsampling operation across the full depth of the network, far beyond the early visual cortex stages where it has biological warrant. Barnard and Casasent established formally in 1990 that this operation discards positional information irreversibly at every layer where it is applied, and repeating it into regions that correspond to V4 and inferotemporal cortex compounds this loss without the compensating transition to qualitatively different computations that the biological hierarchy performs. Third, the pooling-as-complex-cell analogy that motivated this design reflects a misreading of what complex cells compute. The spatiotemporal energy model formalizes complex cell behavior as geometry extraction: detecting the presence and orientation of a local edge structure robustly, abstracting over photometric accidents of contrast polarity and sub-wavelength phase that are not geometrically meaningful. Pooling is a tolerable first-stage approximation of this behavior, but as a general-purpose invariance mechanism repeated across the full depth of the network, it is attempting something categorically different, namely object-level position invariance through spatial subsampling, which achieves its goal by discarding exactly the geometric information that the energy model preserves. Treating pooling as a scalable, indefinitely repeatable implementation of complex cell behavior—rather than as a first-stage approximation with a natural biological endpoint at V3—conflates two operations that differ not in degree but in kind, and crucially it removed the principled criterion for confining the S-C operation to early visual cortex: because pooling was understood as a general-purpose invariance mechanism, the field had no architectural reason to stop repeating it. We survey how capsule networks, group-equivariant CNNs, PDE-based networks, and vision transformers each address one or two of these limitations while leaving the others intact. We propose six desiderata that a more biologically complete architecture would need to satisfy and argue that satisfying them requires treating the visual cortex’s solution as a coherent package in which each component depends on the others working correctly, rather than as a menu of independently selectable principles. Full article

Surface engineering strongly influences the performance, reliability, and safety of medical and biomedical devices, yet failures often originate at interfaces rather than in bulk materials alone. This review addresses the fragmented evidence base linking coating selection, interphase design, qualification testing, advanced characterization, and data-driven durability analysis. The objective is to provide an integrative, failure-mode-based framework for implants, reusable instruments, inhalation systems, diagnostics, wearables, and implantable electronics. A narrative synthesis of the peer-reviewed literature in coatings, biomaterials, electrochemistry, reliability, standards, and materials informatics was conducted, with qualitative tables used only when protocols were too heterogeneous for numerical pooling. The review compares physical vapor deposition (PVD), chemical and plasma-enhanced chemical vapor deposition (CVD/PECVD), atomic layer deposition (ALD), sol–gel/organically modified silica (ORMOSIL) hybrids, plasma polymers, parylene, bioactive or antimicrobial surfaces, and electronic encapsulation strategies. The main finding is that no universally superior coating exists; reliable performance depends on matching architecture and characterization to the dominant failure pathway, substrate compliance, geometry, sterilization or physiologic exposure, and the standards-constrained endpoint. The review further shows how electrochemical diagnostics, interfacial mechanics, multiphysics models, survival/reliability statistics, and carefully governed AI workflows can be combined to support service-life prediction and decision-oriented qualification. Full article

We develop a collection of nontrivial examples that illustrate and test recent stability results for linear control systems ($LCS$) on Lie groups. We treat the main structural classes: Abelian (${\mathbb{R}}^n$), nilpotent (Heisenberg), solvable non-nilpotent (rigid motions of the plane $SE\left(2\right)$), compact semisimple ($SO\left(3\right)$), noncompact semisimple ($SL(2,\mathbb{R})$ via Iwasawa decomposition) and mixed/Levi-type groups. The examples are designed to (i) show the sharpness of geometric boundedness criteria, (ii) exhibit typical failure modes (exponential escape, polynomial central drift, noncompact neutrals), and (iii) demonstrate how the canonical quotient and suitable outputs recover $BIBO$ stability. The executive framework ($ICS$ existence/uniqueness, canonical quotient $G/\Gamma$, $BIBO$ characterization, robustness and ISS-type bounds) is briefly recalled; the main part of the paper consists of detailed worked examples implementing the practical checklist for applying these theorems. Full article

Autonomous driving systems (ADSs) are reliable only when heterogeneous sensors, estimation algorithms, and safety mechanisms are engineered as a single coherent safety-critical measurement system rather than as loosely coupled modules. Production stacks integrate cameras, LiDAR, automotive radar, and GNSS/IMU, yet deployment remains constrained by modality-specific failure modes, calibration and synchronization drift, and out-of-distribution (OOD) conditions that violate modeling assumptions. These limitations induce overconfidence and downstream decision errors whenever planning assumes certainty sharper than sensing can justify. This survey introduces a sensor-centric framework linking measurement physics, uncertainty propagation, fusion integrity, safety assurance, and risk-aware planning and control. We formalize what each modality physically measures; unify probabilistic, evidential, and conformal uncertainty representations; analyze filtering, factor-graph, BEV, transformer, and state-space fusion architectures with an emphasis on robustness and graceful degradation; and generalize aviation-style integrity concepts (RAIM/ARAIM) to multi-modal autonomy. The distinctive contribution is a single sensor-to-assurance throughline in which every uncertainty representation is tied to its measurement physics, every fusion architecture is evaluated against an explicit integrity-monitoring requirement generalized from RAIM/ARAIM, and every safety-standard clause is mapped to a concrete architectural mechanism. We map these mechanisms onto ISO 26262, ISO 21448 (SOTIF), ISO/PAS 8800, ANSI/UL 4600, and the UNECE framework, and connect perception uncertainty to decision-making through chance-constrained MPC and formal safety filters (RSS, CBF). Industry case studies and emerging V2X and generative-simulation approaches close the loop to deployable safety arguments. Full article

This study investigated the relationship between the surface microstructure of pearlite steel wheels and the formation of fatigue cracks during the braking process by using a full-size wheel braking test rig. After fatigue failure, the surface microstructural evolution and fatigue crack initiation and propagation of the wheel sample were systematically analyzed by optical microscope (OM), scanning electron microscope (SEM), and transmission electron microscope (TEM). The results showed that after braking of 1572 cycles, a large number of fatigue cracks formed at the wheel tread, which caused the wheel to break. After fatigue failure, some dark areas formed at the wheel tread, which were composed of Fe₃O₄ compounds. This indicates that severe oxidation was produced at the wheel tread during braking due to the high temperature. After fatigue failure, a continuous thermal white etching layer (T-WEL) was formed in some areas of the wheel tread, while crescent-shaped T-WEL was found in other areas. The microstructure of the T-WEL was composed of martensite phase. The rapid increase and decrease in temperature at the wheel tread during the braking process caused martensitic transformation at the wheel tread. The hardness of the sample reached to about 900 HV in WEL and it reduced with the increase in distance from the surface. The cracks were initiated from the surface and gradually propagated into the matrix. However, the crack propagation mode in the continuous T-WEL and crescent-shaped T-WEL was different. In the continuous T-WEL, the continuous T-WEL of the wheel can be peeled off during the braking wear process, and then the crack was gradually propagated into the matrix in the T-WEL peeled area. As for the crescent-shaped T-WEL, due to the large hardness difference between T-WEL and pearlite, the crack initiated at the interface between the T-WEL and pearlite and gradually propagated into the matrix. Full article

Table olive processing comprises multiple stages in which physical, chemical, and biological hazards may occur. Although risk assessment is a core element of Hazard Analysis and Critical Control Points (HACCP) systems, the selection of assessment tools remains insufficiently standardized. This study compared a 4 × 4 risk matrix and Failure Mode and Effects Analysis (FMEA) for hazard evaluation in Spanish-style and Californian-style table olive processing. Hazards were assessed across 41 processing stages for Spanish-style olives and selected key stages for Californian-style olives using probability × severity in the 4 × 4 matrix and severity × occurrence × detection in FMEA. Significant hazards were further evaluated using the Codex Alimentarius decision tree to identify critical control points (CCPs) and strengthened prerequisite programs (PRPs). Both tools identified similar significant hazards, including biological hazards associated with fermentation, brine management, storage, container sealing, and heat treatment, as well as physical hazards from foreign bodies and chemical hazards related to heavy metals, pesticide residues, mycotoxins, and food-contact material migration. FMEA provided greater analytical detail through the detection parameter, whereas the 4 × 4 matrix was simpler and more practical for complex flow diagrams. Overall, both tools were suitable for HACCP-based risk assessment in table olive processing. Full article

Landslides are the most common geological hazards, and chemical reinforcement is an effective method for enhancing the stability of soil slopes. Based on the coupled Eulerian–Lagrangian method, finite element analyses were conducted to develop a simple design approach for soil slope reinforcement using the curing agent. First, the effects of internal friction angle, cohesion, soil unit weight, slope height and angle on the slope stability were systematically quantified through 93 numerical cases. On this basis, an empirical formula was established for the factor of safety (FOS) of soil slope, and a method for determining the failure mode was proposed using a dimensionless parameter and two critical values related to slope angle. Subsequently, the reinforcement performance of the SH curing agent was investigated by varying the reinforcement position and length. The results indicate that the reinforcement of Case I-II-III and Case I-II provide the best performance, and the optimum reinforcement length was determined for different slope conditions. For slope angles ranging from 25° to 65°, the FOS after reinforcement was found to increase by 12.1% to 18.8% compared with that before reinforcement. Based on the FE results, empirical formulae for predicting the FOS of reinforced slope were further developed. Finally, a simple design approach was proposed for soil slope reinforcement with curing agent. The proposed method provides a convenient and effective reference for engineering practice in soil slope reinforcement with curing agents. Full article

The extreme thermal environment during the underground coal gasification (UCG) process poses a severe threat to the stability of the gasification cavity and the integrity of the surrounding rock. This paper aims to reveal the thermo-mechanical response characteristics and damage evolution mechanism of sandstone under true triaxial unloading conditions following exposure to high temperatures. Sandstone specimens were thermally pre-treated at five temperature gradients (25 °C, 200 °C, 400 °C, 600 °C, and 800 °C) and subsequently subjected to true triaxial loading and unloading experiments. The effects of varying temperatures on the strength, deformation parameters, dilation angle evolution, and macroscopic failure modes of the sandstone were systematically analyzed. The results indicate a significant critical transition point in the mechanical behavior of the sandstone at 400 °C. Below this threshold, thermal-induced microcrack closure leads to an increase in peak strength (with the peak strength at 800 °C increasing by approximately 67% compared to room temperature). Conversely, above 400 °C, thermal damage to the mineral grains intensifies, causing the crack propagation pattern to transition from brittle shear to a complex tension-shear splitting mode, accompanied by severe dilatancy (with a generalized Poisson’s ratio exceeding 0.8). Based on these findings, this study proposes a stage-wise damage evolution model alongside a targeted zonal support strategy, recommending the application of high-prestressed support in high-temperature zones above 400 °C to suppress tensile failure. Ultimately, this research provides a crucial theoretical basis for evaluating the long-term stability of high-temperature underground engineering projects and ensuring operational safety. Full article

Grouting and sealing in gas drainage boreholes are two of the critical measures to ensure efficient coal seam gas extraction. However, traditional cement grouting often leads to debonding and cracking of the slurry–coal cemented body under external load, resulting in poor sealing performance. To suppress crack propagation and achieve borehole reinforcement and efficient sealing, this study compares the mechanical properties and crack evolution characteristics of slurry–coal cemented samples grouted with different modified materials. Five types of cement-based sealing materials, including ordinary Portland cement, were used for grouting coal rock in boreholes. By employing an acoustic emission signal acquisition system and a non-contact full-field strain measurement system, the tensile mechanical properties of coal before and after grouting were compared. The influence of material properties on the reinforcement capacity of borehole coal was analyzed, along with the failure process characteristics and final failure morphology of the slurry–coal cemented body under Brazilian splitting load. Finally, the effects of material toughness and bond strength on the brittleness index and failure mode of the slurry–coal cemented samples under Brazilian splitting conditions were discussed. The results show that the tensile strength improvement rates of the samples were 26.9%, 55.3%, 48.4%, 8.6%, and 45.6%, respectively. Distinct from previous studies focusing on fractured grouting or intact coal rock, this work for the first time systematically reveals the non-monotonic influence of the combination of material toughness and bond strength on the reinforcement effect of borehole coal samples and proposes an evaluation framework based on quantitative acoustic emission crack type analysis and the concept of effectiveness threshold. The varying degrees of tensile strength enhancement indicate differences in the reinforcement capabilities of grouting materials with different properties. The acoustic emission signals during the failure process of the slurry–coal cemented body exhibited typical stage-specific characteristics, though material properties altered the failure modes. By quantifying the intrinsic properties and crack characteristics of the slurry–coal cemented body using the brittleness index and grayscale histograms, this study provides a theoretical basis for guiding efficient sealing of gas drainage boreholes through an in-depth understanding of the mechanical behavior and crack evolution of borehole coal samples before and after grouting under Brazilian splitting conditions. Full article

Short circuits and subsequent fires resulting from objects impacting the bottom of vehicle power battery packs considerably jeopardize electric vehicle (EV) operations. This study investigated underbody impacts in EVs and the overall mechanical properties of battery cells. Key features of road debris were extracted and simplified to establish a geometric parameter structure model and determine realistic battery pack responses to debris impact. Quasi-static compression and dynamic impact tests on a prismatic lithium-ion battery (LIB) and power battery pack followed. Macroscopic mechanical responses, deformation failure modes, and internal jellyroll damage of cells and packs were evaluated, and constitutive equations and failure parameters were derived to develop a finite element model, whose effectiveness and reliability were verified by comparing simulation results with experimental data. Finally, a homogenized model of the prismatic LIB and power battery pack was constructed, which effectively predicted the macroscopic mechanical response and internal short-circuit failure under mechanical loading. However, simulation and test results revealed certain deviations in cell indentations under battery pack bottom impacts, presumably because the FEMs neglect the dynamic strain rate effects of electrolyte and cooling liquid. Overall, this study elucidates safety risks to cells and their key components under power battery pack bottom impacts. Full article

Aluminum is undoubtedly a key material in modern industry. Flat plates made of aluminum alloys are widely used in construction, aeronautics, automotive, and others. The current paper presents an analysis of the behavior of a thin plate made of Al7075-T651 aluminum alloy, subjected to a uniaxial stress, and clamped at one end. The results of the numerical simulation with FRANC2D software have been used for accurate determination of the stress intensity factors (K_I, K_II) and being validated for the simple cases using analytical calculations. The Failure Assessment Diagram (FAD) based on the toughness ratio Kr and the load ratio Lr has been used to evaluate the structural integrity of cracked components based on the load, its position, crack size, and the fracture properties of the material. The FAD analysis results highlight the significant influence of crack position on the values of the K factor. The edge and inclined cracks lead to increases in stress intensity factors and to the occurrence of mixed-mode loading conditions. The study demonstrates the effectiveness and usefulness of the proposed methodology in the analysis of structures with discontinuities and emphasizes the importance of crack positioning in assessing the safety of engineering components. Full article

This study investigates the shear damage mechanisms in reinforced concrete (RC) beams through non-linear numerical modeling. Using the Finite Element Method (FEM) in ABAQUS, a Concrete Damaged Plasticity (CDP) framework was validated against experimental results and subsequently utilized for a 36-model parametric investigation. The study isolated the influence of stirrup spacing, diameter, and yield strength to evaluate their roles in ultimate shear capacity. The results indicated that while increasing stirrup diameter yielded modest capacity enhancements of approximately 7%, the impact of increasing yield strength was negligible, as the failure modes were primarily governed by concrete web crushing before reinforcement yielding could occur. These physical limit states were compared against the linear Truss Analogy adopted by major design standards—including ACI 318-19, Eurocode 2, and TS 500—to quantify discrepancies in heavily reinforced sections. The findings reveal that, strictly within the investigated parameter space (a/d = 2.67, f’_c = 28.5 MPa), current linear equations can significantly overestimate the physical capacity gains provided by reinforcement modifications. These observations are configuration-specific and highlight the need for cautious application of linear models in heavily reinforced scenarios. Furthermore, the study suggests that utilizing 3D beam elements for transverse reinforcement provides a more nuanced representation of shear transfer mechanisms, such as dowel action, compared to standard truss models. Full article

Objectives: This study aimed to compare the acute effects of high-intensity resistance training (HIRT), low-intensity resistance training (LIRT), and low-intensity resistance training with blood flow restriction (LIRT-BFR) on heart rate variability (HRV), rating of perceived exertion (RPE), total load (kg), and number of repetitions in young trained men. Methods: Thirteen volunteers (21.5 ± 1.6 years; 178.2 ± 8.0 cm; 75.7 ± 8.0 kg) performed three training sessions with six upper- and lower-limb exercises in repetition-to-failure mode. HIRT was performed at 70% 1RM, four sets and 90 s of rest; LIRT at 30% 1RM, four sets and 30 s of rest; and LIRT-BFR at 30% 1RM, four sets, 30 s of rest, and cuff pressure at 80 mmHg. The rest interval between training sessions was 72 h. Results: Total load was higher during LIRT compared with LIRT-BFR (p < 0.05), with no significant difference compared with HIRT (p > 0.05). The number of repetitions was greater in LIRT than in HIRT (p < 0.05), with no significant difference compared with LIRT-BFR (p > 0.05). RPE was lower in LIRT compared with HIRT and LIRT-BFR (p < 0.05). Time-domain parameters SDNN significantly decreased across all protocols (p < 0.001), whereas RMSSD showed no differences. Frequency-domain components (LFnu, HFnu, and LF/HF) showed no significant differences. Conclusions: LIRT elicited lower perceived exertion compared with HIRT and LIRT-BFR and higher repetition performance, whereas LIRT-BFR, despite showing similar autonomic responses, produced greater perceptual stress, resembling that of HIRT. Full article

To investigate the influence of dry connection methods on the seismic behavior of assembled composite walls, four assembled composite walls were designed and tested. Various dry connection techniques were adopted for the horizontal interfaces, namely sleeve grouting connection, welding connection, box connection, and bolted connection. The failure process, failure mode, bearing capacity, rigidity, steel bar strain, and energy absorption performance of the specimens were investigated through quasi-static cyclic loading tests. The results indicate that all types of connectors can effectively transfer loads and satisfy the conceptual design principle of “strong joint and weak component”. The damage evolution of the specimens is essentially identical, and the limiting drift angles all exceed 1/90. In addition, the shear resistance of the specimens with different connection methods is preliminarily analyzed and estimated. Full article

To meet the increasing demands of aircraft engines for high thrust-to-weight ratios and elevated turbine inlet temperatures, turbine blades have been progressively designed with thin-walled structures. It has been demonstrated that the mechanical properties of Ni₃Al-based single-crystal alloys (SXs) are highly sensitive to specimen thickness. In this study, the high-cycle fatigue behavior of [111]-oriented Ni₃Al-based SXs with wall thicknesses of 0.3, 0.5, and 0.8 mm was systematically investigated under tensile–tensile loading conditions at 1000 °C. The results revealed that, as the wall thickness decreased, the fatigue life of the alloy significantly deteriorated, while the crack initiation site gradually shifted from the specimen interior toward the surface and near-surface regions. Furthermore, the fatigue failure mode transitioned from being dominated by internal defects to being controlled primarily by near-surface damage. Near-surface damage induced by high-temperature oxidation and geometric constraints was identified as the primary factor responsible for the degradation of the high-cycle fatigue performance of the SXs. In addition, the cyclic deformation behavior at 1000 °C was governed by the synergistic effects of dislocation climb, cross-slip, and γ′-phase shearing. This study provides both theoretical guidance and experimental evidence for the structural optimization of next-generation single-crystal turbine blades for advanced aircraft engines. Full article