Search Results (8,162)

In-line cavitation is relevant to many continuous processes; however, its intensity depends on flow rate, available pressure, temperature, fluid properties, and plant conditions, complicating the maintenance of a repeatable regime within a prescribed band. This paper presents the DVRA, an actuated Venturi module with a Reuleaux triangular cross-section for in-operation regulation of hydrodynamic cavitation through device configuration. The novelty lies in combining two degrees of freedom—an in-operation adjustable hydraulic throat and boundary-imposed swirl forcing—within a compact in-line device: all rotation is confined to the module, and no rotation of the process line is required. The hydraulic throat is tuned via an actuated elastomeric liner, while swirl is generated by external end collars. Reproducible operational conventions are introduced together with a normalized input set and a configuration-space formalism that distinguishes admissible from achievable configurations. Regulation is cast as a control-oriented inverse mapping given a target band for an in-line estimated cavitation indicator and standard industrial measurements of flow rate, pressure, and temperature; configuration commands are selected to keep the indicator within bounds. The contribution is methodological and provides an implementable basis; comprehensive validation and performance benchmarking are outside the scope of this paper and will be reported separately. Full article

The main aim of this study was to evaluate the applicability of a low-cost, double-wall gas metal arc welding (GMAW)-based wire arc additive manufacturing (WAAM) process for aluminum alloy AlMg5, with an emphasis on microstructural heterogeneity, layer-dependent defect formation, and their implications for mechanical performance and geometric characteristics. A Taguchi L9 (3³) design of experiments was employed to investigate the influence of welding current (40–60 A), shielding gas flow (10–20 L/min), and arc correction (0–40%) on wall geometry, material utilization, and overall process quality through multi-response optimization. The optimal parameter set (60 A, 15 L/min, 0% arc correction) resulted in a 54.9% improvement in the Grey Relational Grade compared to the lowest-performing configuration. Metallographic analysis revealed heterogeneous grain evolution governed by the multilayer thermal history, with porosity levels ranging from 3.20% to 3.49% and lack-of-fusion defects preferentially concentrated in interlayer and mid-height regions. The fabricated high-wall structure exhibited hardness values between 72 and 85 HV and an average ultimate tensile strength of 175 MPa. The observed mechanical scatter was consistent with localized microstructural heterogeneity and spatial defect distribution. The results demonstrate that geometric evaluation alone is insufficient as a quality metric for WAAM components and must be complemented by metallographic integrity assessment. Overall, the study highlights the importance of direct parameter optimization in double-wall WAAM structures to mitigate defect formation and enhance mechanical reliability under industrially accessible deposition conditions. Full article

Facial emotion recognition (FER) systems can serve as a valuable non-invasive tool for assessing emotional states linked to mental health. However, two main issues hinder their adoption in clinical settings, including privacy concerns inherent to centralized data processing and the lack of transparent decision-making processes. This paper proposes a privacy-preserving and explainable FER framework that implements a collaborative distributed training approach for lightweight convolutional neural network (CNN) architectures across three heterogeneous datasets: RAF-DB, ExpW, and FER2013. SHapley Additive exPlanations (SHAP) guided the optimization of CNN filter configurations, prioritizing high accuracy, cross-dataset generalization, and interpretable, trustworthy explanations. A multidimensional explainability evaluation framework is developed that combines perturbation-based faithfulness and feature localization metrics into a Global Explanation Quality Score (GEQS) for quantitative assessment of explanation quality. A qualitative user study was conducted to assess alignment between human perception and quantitative explainability metrics. Guided by explainability-driven evaluation, a lightweight CNN achieves 74.3% mean accuracy, highlighting its effectiveness for cross-dataset generalization. Results demonstrate alignment between quantitative GEQS and human evaluation, with both identifying the same model architecture. However, qualitative analysis shows that highlighting emotion-relevant facial features does not always ensure user trust. The practical viability of the proposed FER system in resource-constrained clinical environments is demonstrated through implementation on Raspberry Pi 4 with integrated SHAP explainability, achieving 60 ms inference time. Full article

The precise control of piezoelectric actuators is limited by inherent hysteresis, creep, and nonlinear behavior, which necessitate high-resolution displacement sensing for effective closed-loop operation. Although optical interferometers can achieve nanometer and sub-nanometer resolution, their practical implementation is often constrained by complex optical alignment, sensitivity to environmental disturbances, and limited robustness in high-speed measurements. Optical triangulation sensors offer a more robust and straightforward alternative; however, their resolution is typically insufficient for nanometer-scale displacement measurements. In this study, a novel optical triangulation sensor based on a two-stage geometric optical amplification scheme is proposed for measuring the expansion of piezoelectric stacks. The method relies purely on geometric optical amplification and does not require interferometric techniques or complex signal processing. Using off-the-shelf optical components and an industrial imaging sensor, the proposed system achieves a displacement resolution of 109.6 nm, a repeatability of 74.62 nm, and an accuracy of 98.81% with a maximum error of 207.14 nm under hysteresis measurements. The achieved resolution is primarily limited by the spatial resolution of the camera sensor, indicating that further improvements are possible through optimization of the optical configuration or the use of higher-resolution imaging devices. Owing to its simplicity and robustness, the proposed sensor is well suited for real-time closed-loop control of piezoelectric actuators. Full article

This study investigates the performance of a multi-media bio-scrubbing system that integrates activated sludge with loofah as a biofilm carrier for the purification of complex pollutants from simulated cooking fumes: oils, Non-Methane Hydrocarbons (NMHCs), PM_2.5/PM₁₀, and Volatile Organic Compounds (VOCs). Compared to conventional carriers like activated carbon, the biodegradable and low-cost loofah, with its hierarchical porous structure and balanced hydrophilic–lipophilic properties, showed enhanced support for microbial colonization (achieving a biomass density of 105 mg/g) and pollutant adsorption. The system achieved high removal efficiencies in lab-scale tests: 97.4% for total VOCs (including 96.5–100% removal of recalcitrant alkanes and olefins), 91.3% for oils, and >88% for PM_2.5/PM₁₀. Mechanistic analysis indicated that the biofilm was dominated by Proteobacteria and Actinomycetes, and the synergistic effect between physical adsorption (via loofah’s porosity) and biodegradation (by microbial consortia) enabled stable performance (maintaining >90% efficiency over a 25-day operation) without observed secondary pollution. The loofah-activated sludge configuration demonstrated improved operational stability and the potential for lower operating costs compared to single-medium systems in this experimental setting. This work explores a promising, eco-friendly approach for treating simulated cooking fumes, utilizing renewable biomass carriers and biological processes, which could contribute to cleaner production strategies. Full article

The pantograph–catenary system is a critical component of the traction power supply network. Due to hard points on the overhead contact line and vibrations of the pantograph, pantograph–catenary separation may occur, leading to offline DC arc events. To investigate the characteristics of DC arcs generated during pantograph–catenary separation in metro systems, this study constructs a laboratory platform that simulates the offline process and analyzes the electrical characteristics, optical intensity, and arc-burn duration under different electrode separation conditions. First, a DC pantograph–catenary offline arc simulation platform is developed using a contact wire, a carbon-strip pantograph slider, and a linear motor, enabling slider movement in both horizontal and vertical directions. Second, offline discharge experiments are conducted to compare the discharge process and electrical arc characteristics with and without horizontal slider motion. Finally, arc luminosity and burn duration are measured under various electrode separation configurations, and the influence of voltage level, current level, and electrode material is examined. Experimental results reveal a significant polarity effect, where the arc burn duration is notably longer when the contact wire serves as the cathode than when the carbon slider serves as the cathode. At the instant of separation, the high electric field intensity within the micro-gap triggers pronounced “peak phenomena” in both arc resistance and power, accompanied by abrupt voltage surges and transient current dips. Furthermore, the introduction of horizontal motion modulates the arcing process, causing the stable arcing voltage to follow a distinctive trend of a slow increase followed by a gradual decrease, which differs from static separation characteristics. Finally, this study demonstrates that voltage levels exert a more dominant influence on arc luminosity and duration than current levels, while the maintenance voltage of the arc channel remains significantly lower than the air breakdown voltage. Full article

Carbon Capture and Sequestration (CCS) is one of the most important technologies that can help many countries to reduce emissions into the atmosphere and lower their carbon footprint, which in turn can help to achieve the net-zero goal. However, when CO₂ is injected into a suitable geological formation in the subsurface during CCS operations, it is essential to ensure that the well integrity of the legacy well within the Area of Review (AoR) is maintained so that the injected CO₂ will not make its way to the shallow formation, which can ultimately contaminate the Underground Source of Drinking Water (USDW) and make it to the surface, which can have harmful effects on the environment and human health. Hence, this paper presents a semi-quantitative risk assessment framework for legacy wells within a CCS Area of Review (AoR) and for an Underground Injection Control (UIC) Class VI injection well. The method converts a Feature Event and Process screening into an interaction matrix (IM), assigns probability × severity scores using an incident potential matrix (IPM), and derives cause-and-effect metrics to rank barrier elements and wells. The legacy wells are evaluated using a 5 × 5 IM (casing, cement sheath, water composition, gas/CO₂, and USDW), and the injector is assessed using a 7 × 7 IM (adding tubing and packer). From the results, it was found that the risk levels of the Types 2 and 3 wells were the highest, while Types 4–6 clustered in the medium-to-low range, and Types 7–9 and the Class VI well were dominated by low/very low classes. Therefore, it was concluded that the level of risk associated with legacy wells in the AoR depends upon well completion, the well configuration, the number of well barriers, and the depth to which the well penetrates. It was further found that, within the multi-barrier well, the risk level of the first barrier is lower; the risk score continues to increase with each subsequent barrier above it. The most critical elements in the given risk assessment framework for legacy wells in the AoR remain the well cement sheath, casing, and USDW. Meanwhile, the components that affect the well are water composition and the presence of gas/CO₂. Full article

An up-to-date catalog of residents space objects orbiting Earth requires a critical balance between computational efficiency and orbital prediction precision. This work presents HSGP4, a hybrid orbit propagator specifically tailored for Galileo-type orbits that enhances the classical SGP4 analytical model using Artificial Neural Networks. The methodology centers on a non-invasive hybridization process that utilizes high-fidelity pseudo-observations to forecast SGP4 error residuals. A core contribution is the introduction of the Hybrid Two-Line Element format, which encapsulates neural model parameters alongside traditional orbital elements, ensuring seamless integration with existing catalog infrastructure. The development process involved comprehensive Exploratory Data Analysis and sensitivity analysis, which identified the argument of latitude as the most influential variable for correcting SGP4 errors in the MEO region. To ensure statistical robustness, a hierarchical selection strategy was implemented. This reduced an exhaustive search space of 32,256 candidate architectures to a final subset of optimized configurations. Validated against a decade of TLE data, the results confirm that HSGP4 effectively captures missing dynamic patterns and significantly improves ephemeris accuracy. By forecasting SGP4 error residuals, this hybrid approach provides a high-fidelity correction layer. It compensates for the limitations of analytical theories without requiring complex numerical integration. Full article

Uncontrolled urban sprawl represents a significant challenge for many countries worldwide. This article analyzes discussions sparked by increased land consumption driven by urban sprawl in Timișoara’s peri-urban area, Romania. In this context, the objective is to identify the processes of transformation and dysfunction of spaces under the effect of peri-urban expansion. Methodologically, geospatial and satellite data were utilized to assess the evolutionary trend of peri-urbanization. Secondly, an evaluation of land-use types at the level of peri-urban sub-neighborhoods was conducted. Furthermore, an analysis of the online reactions of the inhabitants of these spaces was conducted. The results demonstrate differentiated urban growth patterns in the spatial expansion dynamics of peri-urban spaces, which have emerged along major communication axes and in response to the configuration of available land and proximity to the urban core of each locality. Based on LCR/PGR indicators, these patterns can be further categorized into compact and expansive growth models. Furthermore, deficiencies in fundamental infrastructure pose a significant challenge for a considerable proportion of the local population. The study provides authorities and urban planners with reflective analyses on how to better manage peri-urban development. The results of the study support coherent, preventive, and sustainable urban development to avoid replicating the dysfunctions observed in the studied areas in other peri-urban areas Full article

Multi-point flexible die (MPFD) exhibits broad application potentials for efficient and controllable forming of curved sheets due to its rapid reconfigurability. Nevertheless, the relatively poor surface accuracy and geometrical accuracy of the fiber-reinforced composite components formed by MPFDs limit the widespread application of this technology. In this study, a novel gas-assisted diaphragm forming (GADF) process based on MPFDs for curved basalt fiber/epoxy resin composite sheets was proposed. The precise control of temperature, pressure and MPFD configuration in the process was realized and verified. The effects of different process parameter configurations on dimple defects and geometrical accuracy were analyzed, and the mechanism of defect generation was investigated. A response surface-based forming accuracy prediction model was developed to analyze the influence of component structural parameters on geometrical accuracy. Based on the predictive model, compensation reconfiguration of MPFDs was carried out to achieve high-accuracy sheet forming. Results demonstrated that increasing pressure exacerbated the dimple while reducing shape accuracy. A moderate temperature (120 °C) was proved optimal for component forming, as both excessively low and high temperatures aggravated dimple and induced geometrical errors. Increasing interpolator thickness effectively reduced dimple defects, but excessive thickness adversely affected component geometrical accuracy. Considering both dimple suppression and geometrical accuracy, the optimal process parameters were determined to be 5 kPa, 120 °C, and 2 mm of interpolator thickness. Through MPFD modification based on the response surface model, the geometrical accuracy of the formed components was improved by 38.85%, achieving high-quality forming of the curved composite sheets. Full article

The electro-Fenton (EF) process is a promising advanced oxidation technology for the removal of micropollutants (MPs) from wastewater. This study aimed to identify energy-efficient operating conditions for H₂O₂ electro-production and EF degradation of the neutral micropollutant carbamazepine (CBZ). The effects of current density, residence time (RT), Reynolds number, pH, and temperature were evaluated, and non-EF removal pathways and flow-configuration effects were quantified. H₂O₂ production was maximized under conditions that sustained current efficiencies ≥ 50%, corresponding to specific energy consumption of 4.0–6.4 kWh/kg H₂O₂. Non-EF removal mechanisms intrinsic to the divided electrolytic cell accounted for approximately 35% of total CBZ removal at an RT of 5 min. Under energy-efficient EF conditions (25:1 H₂O₂:Fe²⁺ and 5 min RT), CBZ removal efficiency reached 95%. Asymmetric flow configurations reduced apparent removal through dilution. In contrast, directing the cathode effluent through the anode enhanced oxidation and reduced treated water volume without additional energy input. Total electrical energy per order of CBZ removal from secondary effluent for EF (73.8–94.4 kWh/m³) was comparable to that of Fenton oxidation (61.1–100.0 kWh/m³). In both processes, H₂O₂ production dominated the energy demand. The results highlight EF as a feasible, energy-competitive option for removing persistent MPs from wastewater effluents. Full article

Background: Accurate forecasting of lung tumor motion is crucial for precise radiotherapy. Deep-learning-based markerless tracking methods have been explored, but extending these approaches to predict future tumor trajectories remains largely unaddressed. We address this by framing markerless lung tumor motion forecasting as a spatio-temporal prediction task using a vision transformer to estimate three-dimensional tumor positions over short horizons. Methods: Digitally reconstructed radiographs (DRRs) generated from four-dimensional computed tomography scans of 12 lung cancer patients were used to train a multi-patient (MP) model. Patient-specific (PS) models trained solely on planning data were compared, and the MP model was further fine-tuned using a small number of patient-specific treatment images under realistic clinical constraints. Models processed sequences of 12 DRRs, with performance evaluated via root mean square error. Results: The results indicate that low-resolution inputs with larger patch sizes outperform higher-resolution configurations by reducing image noise. PS models require extensive data to match MP performance, whereas fine-tuning the MP model with limited patient-specific data achieves comparable or superior forecasting accuracy at a lower cost. Conclusions: These findings demonstrate that Vision Transformers can extend markerless tracking methods to accurate short-term forecasting and highlight fine-tuning as an efficient strategy for personalized prediction. Full article

Music source separation (MSS) focuses on decomposing a mixed audio signal into individual instrumental components and is increasingly relevant for music production, restoration, remixing, education, and music information retrieval. Deep learning methods, particularly U-Net architectures operating on time–frequency representations, have recently advanced the state of the art beyond traditional signal-processing techniques. This work presents an optimized multi-source U-Net model for separating selected musical instruments from stereo mixtures. The system uses magnitude spectrograms generated by the short-time Fourier transform and is trained and evaluated on the MUSDB18 dataset. We systematically examine architectural and training-related factors, including normalization strategies, dropout placement, optimizer selection, loss weighting, data augmentation, and spectrogram-domain modifications. Separation quality is measured using BSS Eval metrics, assessing artifacts, interference, and distortion. Experimental results show that the proposed configuration achieves competitive performance relative to established convolutional and U-Net-based open-source systems, especially in terms of vocal track separation, offering practical insights into designing efficient models for multi-instrument separation. Full article

This paper reports an advanced statistical physics modeling to elucidate the adsorption process of two relevant pharmaceuticals, namely amoxicillin (AMOX) and imipramine (IMP), on aluminum-pillared clay. A double-layer model was used to interpret the adsorption mechanism of these pharmaceuticals at 298–318. This model indicated the presence of molecular aggregation for IMP adsorption via the formation of dimers or trimers, and a monomolecular AMOX separation process. The removal of both compounds was endothermic, showing better adsorption capacity for IMP (82 mg/g) than for AMOX (37 mg/g). Calculated adsorption energies (ΔE₁, ΔE₂ < 40 kJ/mol) confirmed a physisorption mechanism, which may be governed by hydrogen bonding due to the adsorbent chemistry and adsorbate molecular structure. Configurational entropy and free enthalpy were calculated to analyze the thermodynamics of AMOX and IMP adsorption equilibria. These thermodynamic functions confirmed the molecular disorder during adsorption and system’s spontaneity. This study contributes with new theoretical findings for unraveling complex adsorption mechanisms, at the molecular level, of pharmaceutical molecules, with the aim of intensifying water depollution systems. Full article

This paper presents FireVision, an innovative platform and model for real-time fire detection and monitoring. The platform utilizes automated drone flights to collect high-resolution imagery in both suburban and forested settings. Ensemble deep learning inference, based on Mask R-CNN weak learners, is employed to trigger alerts. Detection performance is further enhanced by integrating ResNet-50, ResNet-101, and ResNet-152 classifiers, which can be deployed in the cloud or on the drone’s edge co-processing units. Additionally, a fire criticality index is introduced, leveraging detection bounds and masks to assess the severity of fire events, alongside an automated drone path-planning algorithm for identifying critical fire incidents. Experiments were conducted using a supervised, mask-annotated dataset to evaluate model accuracy and inference speed across various cloud and edge computing configurations. Results indicate that ResNet-101 surpasses ResNet-50 by 5 to 12.5 percent in mAP@0.5 mask accuracy, with an 18 percent increase in inference time on the cloud and a 27 percent increase on the drone edge device GPU. In comparison, ResNet-152 achieves a 0.5 to 1.2 percent improvement in mAP@0.5 over ResNet-101, but its inference time is nine times slower in the cloud and 1.3 times slower on the GPU. Full article