Existing railway obstacle detection methods primarily focus on object localization and classification, while lacking the capability to directly support operational risk assessment. To address this limitation, this paper proposes a risk-aware railway obstacle detection framework that integrates track segmentation and lateral distance estimation. The proposed framework jointly optimizes obstacle detection and track segmentation through shared feature learning, and estimates the lateral distance between detected obstacles and track boundaries based on the extracted track geometry. Railway clearance constraints are further incorporated for obstacle risk-level determination, thereby establishing an end-to-end pipeline from obstacle detection to operational risk assessment. Consequently, an end-to-end pipeline is established to classify obstacle risk levels in railway environments. To validate the proposed framework, a dedicated RN-rail-Object dataset is constructed, and comprehensive ablation studies, comparative experiments, and edge deployment evaluations are conducted. Experimental results demonstrate that the proposed method achieves 89.2% mAP for obstacle detection, 93.7% mIoU and 97.1% Dice for track segmentation, while maintaining an inference speed of 214.6 FPS on the RTX A4000 platform. Furthermore, edge deployment demonstrates the feasibility of integrating obstacle detection, track extraction, and geometry-based risk assessment on an edge computing platform, with representative scenarios illustrating the identification of different obstacle risk states. The proposed framework provides an interpretable geometry-based approach for extending conventional obstacle detection toward railway risk-aware perception, showing potential for intelligent railway inspection, early warning, and safety-oriented operation and maintenance.
High-modulus asphalt binders (HMABs) are critical for heavy-duty pavements but are often constrained by low-temperature brittleness and poor dispersion of conventional granular modifiers. This study develops a novel micro-pulverized composite modifier (ZY) integrating hard asphalt, an ethylene–propylene copolymer, and a plasticizer system via high-shear melt blending and centrifugal atomization. The modification effects and underlying mechanisms were systematically investigated through rheological characterization (DSR, BBR, LAS), thermodynamic analysis (DSC), chemical functional group evaluation (FTIR), and microscopic morphological observation (FM), with two commercial high-modulus agents (PR and JK) as benchmarks. The results demonstrate that ZY significantly enhances high-temperature deformation resistance, elevating the Performance Grade from PG 64-22 to PG 82-10, with the complex modulus (G*) consistently exceeding those of PR and JK across the entire temperature sweep range (46–82 °C). At low temperatures, the synergistic toughening effect of the elastomeric copolymer and plasticizer enables a creep stiffness S of 240 MPa and an m-value of 0.298 at −12 °C, satisfying Superpave requirements and ensuring superior stress relaxation capability. The LAS test reveals a fatigue life of 245,000 cycles at 2.5% strain level, representing an approximately 60% improvement over the JK-modified binder. Microscopic characterization (DSC, FTIR, and FM) confirms that the modification mechanism is dominated by physical blending, forming a highly uniform micro-scale multiphase dispersion system: the hard asphalt component integrates into the matrix to achieve viscosity enhancement and stiffening, while the elastomeric copolymer forms finely dispersed spherical microspheres that effectively impede crack propagation and dissipate strain energy. This synergistic design achieves a favorable balance between high-temperature modulus and low-temperature flexibility, offering a promising solution for durable and rut-resistant pavement applications.
Road pavement construction generates a significant initial carbon footprint, mainly related to material production, transport and construction processes. This study comparatively assesses the “cradle-to-site” carbon footprint of selected road pavement structures, with the aim of supporting low-carbon pavement design choices. The analysis was carried out using the Life Cycle Assessment (LCA) methodology, modelled with the open-source software OpenLCA®, adopting a “cradle-to-site” approach limited to the production and construction phases. Several scenarios were examined by varying the percentages of recycled material content (RAP—Reclaimed Asphalt Pavement) used in asphalt mixtures, the functional road category and the pavement structural type. The results show that the reference flexible pavements produce approximately 70 kg CO2 eq/m2, with asphalt mixtures accounting for the largest share. The use of RAP enables emission reductions of up to 14%, although marginal benefits decrease as the recycled material increases. Impacts also vary according to the functional road category, ranging from approximately 50 to more than 90 kg CO2 eq/m2 when moving from local roads to motorways. Within the adopted cradle-to-site boundary, the LCA model supports the early-stage comparison of alternative pavement designs.
Accurate and scalable driver performance assessment is critical for improving road safety and reducing traffic-related injuries and fatalities, particularly in low- and middle-income countries where the majority of global road deaths occur. This paper presents an exploratory proof-of-concept framework for automated driver evaluation that combines real-world dashcam footage, YOLOv8-based object detection, and multimodal large language models (MLLMs), specifically Gemini 1.5 Flash. Two prompting strategies, narrative and rule-based, were designed to assess driver behavior against standardized licensing criteria derived from the California Department of Motor Vehicles (DMV) driving performance evaluation score sheet. The framework was evaluated across 11 manually curated driving scenarios covering intersections, pedestrian crossings, stop signs, cyclists, and emergency vehicles. Ground-truth labels were established through consensus between two traffic engineering experts cross-referencing official California DMV evaluation criteria. In this preliminary evaluation, the rule-based prompt achieved higher agreement with ground-truth assessments (10/11 scenarios, 90.9%) compared to the narrative prompt (7/11 scenarios, 63.6%), particularly in detecting clear rule violations. The narrative approach demonstrated greater contextual flexibility in ambiguous situations. These results should be interpreted as preliminary, given the small sample size, manually curated dataset, and absence of large-scale statistical validation. Nonetheless, the findings illustrate how combining visual detection with structured language-model prompting may support interpretable, policy-aligned driver evaluation. Key limitations include dependence on video quality, limited scenario diversity, absence of temporal behavioral modeling, and reproducibility constraints tied to proprietary API behavior. Future work should expand validation to larger annotated datasets, incorporate temporal sequence modeling, and explore region-specific regulatory adaptation.
Existing comparative studies on metallic mild steel dampers (SDs) and fluid viscous dampers (FVDs) are primarily limited by the coupling of device type with layout variations, the lack of a unified performance metric, and the absence of multi-level evidence under fixed structural configurations. This study overcomes these limitations by comparing SDs and FVDs under strictly identical conditions—same RC frame, same 26 damper locations, same ground motions, and a unified code-specified drift target—across frequent, design-basis, and rare earthquake levels, supplemented by energy dissipation and added damping ratio analyses. Under frequent earthquakes (FEs), the FVD achieves a maximum story-shear reduction of 33% and effectively controls inter-story drift through its velocity-dependent energy-dissipation mechanism. Under rare earthquakes (REs), the SD demonstrates superior performance, providing a 35% maximum story-shear reduction, while maintaining inter-story drift ratios within code-specified limits, owing to its combined stiffness and damping contributions. In terms of energy dissipation, the total cumulative energy dissipated by FVDs is 39.4–67.6% higher than that of SDs under the same ground motions, with added damping ratios averaging 2.42% for FVDs and 2.86% for SDs. These findings suggest that FVDs are more favorable for serviceability and frequent seismic performance, while SDs exhibit better response reduction effects under rare earthquake excitations.
Pile foundations transmit structural loads to deeper subsoil strata whenever the soil in the vicinity of the ground surface lacks sufficient strength and stiffness to ensure an adequate factor of safety against ultimate failure or warrant settlements to remain below acceptable limits. In many in situ conditions, piles are embedded in layered subsoil medium. In several circumstances, piles are subjected to tensile loading. Large and high-rise civil infrastructure subjected to wind loading, transport infrastructure under horizontal loading due to moving vehicles, offshore structures withstanding wind and wave loading, underground structures subjected to hydrostatic pressure due to buoyancy, etc., are some examples where tension loads are imparted on the supporting piles. The imparted uplift loads in these tension piles are balanced by the negative skin friction induced at the pile–soil interface. In this paper, an analytical model using systematic application of established upper bound shear stress theory to three-layered soil configurations has been developed to formulate the ultimate uplift capacity of tension piles in three-layered soil. The model adopted appropriate correlations for upper bound interface shear stresses in different soils as well as tensile failure of pile material itself. The developed solution was validated by comparing with available experimental results. Thereafter, a case study was performed to study the influence of the variation of pile geometries and relative stiffness on ultimate uplift capacities. Important conclusions were drawn from the entire study.
Carbon-fiber-reinforced polymer (CFRP) reinforcement is a potential alternative to steel in corrosive environments. However, CFRP is elastic without ductility, and the seismic performance of CFRP-reinforced concrete (RC) columns is inadequately understood. This study characterizes the intrinsic seismic response of concrete columns reinforced with CFRP cable-type bars and spirals under recorded near-fault ground motions. Three reference steel-RC columns are designed as seismic-resistant, non-seismic-resistant, and with post-cracking stiffness equivalent to the CFRP-RC column. The CFRP-RC and seismic-resistant steel-RC columns were tested under cyclic loading, and the results validated finite element (FE) models. Validated models simulated four columns under cyclic loading, and under 11 near-fault records matched to a capacity-derived elastic target spectrum. The results, bounded by selected ground motions and material constitutive models, show that: (1) The tested CFRP-RC column dissipated about 50% less energy than the steel-RC reference; (2) No material-level failure criterion was met under the suite, although peak base shears exceeded the nominal quasi-static capacities; (3) The CFRP-RC column developed the largest transient drift but minimal residual drift, whereas the steel-RC columns limited transient amplitude via hysteretic dissipation yet accumulated permanent offsets; (4) Response of the CFRP-RC column depends on ground motion energy delivery characteristics: concentration, symmetry, and duration.
Double-layer barrel vault roofs with double-layer vertical walls are widely used in important public buildings because of their high structural efficiency, favorable stiffness-to-weight ratio, and architectural versatility. Although incremental dynamic analysis (IDA) is a widely accepted approach for seismic assessment, it requires numerous nonlinear time-history analyses (NTHAs), resulting in high computational cost. This study presents an artificial neural network (ANN)-based surrogate modeling framework to accurately predict the seismic response of these structural systems, reducing the need for repeated NTHAs, enabling rapid estimation of structural dynamic responses, and facilitating direct development of seismic fragility curves. The proposed framework substantially decreases computational effort while maintaining an effective balance between accuracy and efficiency. A comprehensive seismic damage database is first generated using finite element (FE) models developed in OpenSees. Fragility curves are then obtained using both the conventional IDA procedure and the proposed ANN-based surrogate approach. Results show that the ANN surrogate accurately predicts the responses of structures subjected to scaled ground motions and effectively captures their nonlinear seismic behavior. Furthermore, the resulting fragility curves closely match those from the conventional IDA method, demonstrating the accuracy, reliability, and efficiency of the proposed framework for rapid seismic assessment of double-layer barrel vault structures with double-layer walls.
Automated pavement distress detection is essential for transportation infrastructure maintenance and road asset management. In practice, however, such detectors often need to run on embedded devices mounted on inspection vehicles, and two challenges hinder real-world deployment: (1) cracks and potholes possess markedly different geometric priors—thin, elongated topology versus blob-like shapes—so a generic backbone tends to under-represent at least one class, and (2) accuracy-oriented detectors are typically too heavy for the embedded GPUs commonly mounted on inspection platforms. To address these issues, this paper proposes a lightweight real-time pavement distress detection network. First, a multi-scale coordinate attention (MSCA) module is embedded in the neck so that long-range row/column-wise dependencies are encoded together with multi-scale local context, which is helpful for slender cracks while remaining computationally efficient. Second, a slender-aware detection head (SADH) couples a 1 × k/k × 1 asymmetric branch with the standard square branch, giving the head an explicit inductive bias for elongated objects. Third, a multi-granularity knowledge distillation (MGKD) scheme is designed, which transfers teacher knowledge from a heavier teacher to the proposed student at three complementary granularities—pixel-level attention-masked features, instance-pair relations, and decoupled class-prior logits—thereby covering the three distinct levels of information that a multi-class dense detector relies on. The network is trained and evaluated on the public RDD2022 benchmark together with a supplementary in-house set of asphalt potholes. Under the fixed-seed, single-run evaluation used in this study, the proposed method achieves an mAP@0.5 of 71.65% on the author-defined test split, which is not directly comparable with evaluations on the official RDD2022 test set, with the comparison restricted to seven representative baselines evaluated under the same protocol, and runs at 72.5 FPS on an NVIDIA Jetson Orin Nano. Although its latency is modestly higher than that of YOLOv8s, it retains real-time inference capability for on-vehicle pavement inspection.
Fully enclosed noise barriers (FENBs) are widely used in high-speed railway systems to mitigate environmental noise; however, the transient aerodynamic loads generated by train passage can induce complex structural responses. The relationship between the spatial–temporal evolution of these aerodynamic loads and the dynamic response of the complete FENB structural system remains insufficiently understood. To address this issue, this study develops a sequential computational fluid dynamics–finite element analysis (CFD–FEA) framework that directly relates the transient pressure evolution during the complete train-passage process to the deformation and stress responses of the principal FENB components. The unsteady aerodynamic field generated by high-speed train passage is simulated using a moving-mesh CFD model, and the resulting time-dependent pressure loads are subsequently applied to a finite-element structural model. Train speeds ranging from 250 to 330 km/h are considered. The results reveal strongly transient and spatially non-uniform pressure distributions inside the FENB, characterized by nose-induced compression, a middle negative-pressure region, and wake-induced pressure fluctuations. Both structural deformation and equivalent stress increase with train speed, and the exit stage produces the most pronounced structural response because of the strong negative-pressure effect. Different structural components exhibit distinct response characteristics, with localized stress concentrations occurring in the glass panels and H-section steel columns. By establishing the correspondence between transient aerodynamic pressure evolution, train-passage stages, and component-level structural responses, this study provides a more comprehensive understanding of the aerodynamic load–structural response mechanism of FENBs and provides a basis for structural design and engineering assessment under increasing train speeds.
Currently in the United States, there are more than 623,000 bridges of which 6.8% are in poor condition. Over 100 million trips are taken across these structurally deficient bridges every day and the bridge-related system rehabilitation need is estimated at $191 billion. In 1990’s, calibration of AASHTO LRFD Bridge Design Code involved extrapolation of distributions to evaluate the mean maximum 75-year live loads. However, this derivation was based on a survey of Ontario trucks with small sample size. In the meantime, the Weigh-in-Motion technology improved, and millions of vehicles are measured in various locations on a continuous basis. The objective of this study is to examine site-specific live load spectra based on WIM database in Alabama. Massive volume of data collected by the Department of Transportation is used to compute moments and shears for spans ranging from 30 ft (9 m) to 200 ft (60 m). Cumulative distribution functions of bias ratios (WIM Truck/HL-93 Loading) were plotted on normal probability paper and extrapolated. The findings show that AASHTO provisions for Strength I Limit State are not representative of vehicles at WIM sites located on interstate highways. The findings confirm the importance of continuous traffic data collection and use of reliability analysis procedures.
The shear behavior of corrugated steel web (CSW) composite girders near intermediate supports differs fundamentally from that of conventional beam segments due to the rigid restraint of concrete cross-beams, yet current design codes incorrectly assume the CSWs resist the entire shear force. This paper presents a refined analytical method, based on a three-beam composite model, that explicitly accounts for the cross-beam constraint effect. By assuming a quadratic parabolic distribution of the additional shear-flow intensity along the constraint zone, a closed-form expression for the effective shear force carried by the CSWs is derived. The proposed method is validated against three-dimensional finite element (FE) simulations and existing experimental data, and further corroborated by a parametric study covering varying structural configurations. The results confirm a shear redistribution mechanism characterized by “CSWs unloading and flange sharing.” At the section nearest to the cross-beam, the CSWs actually carry only 65.41% of the total shear force, while the flanges share approximately 35%. In contrast, the conventional code method, which neglects flange shear contribution, severely overestimates CSWs’ shear stress, producing an error as high as 35.29% at the section adjacent to the cross-beam. These findings demonstrate that the cross-beam constraint must be considered in shear design, especially near supports, and the proposed method offers a rational and accurate alternative to existing code provisions.
Connections between earth–rockfill and concrete dams are critical components of hybrid-dam seepage-control systems because material-stiffness contrasts and complex foundation conditions can create localized preferential seepage paths. Using HS Reservoir as a case study, this predictive design-stage assessment employed a full-domain three-dimensional model of the dam–foundation–abutment system and a local three-dimensional model of the cutoff-spur-wall connection. The seepage field, hydraulic gradients, and zonal seepage discharges were evaluated under the normal pool, design flood, and check flood levels, together with the responses of the connection interface and right-abutment grout curtain. Across the three baseline scenarios, the impervious core accounted for 82.2–83.6% of the total head difference at the maximum riverbed section, and the reported control-location gradients remained below the corresponding design values. At the check flood level, the modeled 178 and 179 m head contours passed above the local curtain crest at elevation 177.5 m, identifying an over-curtain seepage pathway. From the design flood level to the check flood level, right-abutment discharge increased from 259.86 to 544.49 m3/d (109.5%), while total discharge increased by 28.6%. Flow in the connection zone diverted around and beneath the cutoff spur wall, and the connection-surface gradients increased with reservoir level. These model predictions characterize the three-dimensional seepage response of the connection zone and right-abutment seepage-control system and can inform curtain-crest review, construction quality control, and post-impoundment monitoring.
This study proposes a quality-aware multimodal framework for health-state assessment and remaining useful life (RUL) prediction of cold-region tunnels. The framework integrates structural-response, environmental, apparent-defect, and engineering-inspectiondata, with the apparent-defect pathway jointly encoding raw images through a convolutional neural network and structured defect variables. Five data-quality dimensions-completeness, accuracy, consistency, timeliness, and traceability are incorporated intoreliability-guided multimodal fusion. Their base weights were re-audited through two rounds of expert consultation, each comprising 323 valid questionnaires. The Cr-weighted group analytic hierarchy process yielded weights of 0.0548, 0.1326, 0.1372, 0.2279, and 0.4474, respectively, with a group consistency ratio of 0.0455; the ranking remained stable under one-at-a-time +10% perturbations. In the primary tunnel case study, the framework achieved 89.7% health-state accuracy, a 6.3% RUL mean absolute percentage error, and 84.1% accuracy under Gaussian perturbation of standardized numerical inputs at a noise scale of 0.15. To further examine the reliability contribution of data-quality information, an independent field panel comprising 600 segment-month observations from 25 segments across three operational tunnels was evaluated using target-excluded specifications, two-way fixed effects, leave-one-tunnel-out validation, multiple baseline models, and five fixed random seeds. A one-standard-deviation increase in lagged quality instability was associated with a 0.0151 increase in the subsequent state-error index (95% CI: 0.0118-0.0184; p < 0.001). In cross-tunnel random-forest tests, incorporating quality information increased mean R2 from 0.8277 to 0.8323 for state-error prediction and from 0.8517 to 0.8673 for RUL-contraction prediction, with both improvements significant in paired tests (p < 0.001). Split-conformal intervals achieved mean cross-tunnel coverage of 95.8% and 95.9%, respectively. These findings demonstrate that data-quality information provides a modest but statistically supported improvement in cross-tunnel reliability, whilethe principal contribution lies in integrating auditable data governance, reliability-aware fusion, and engineering decision support within a unified tunnel health-management framework.
The shaft resistance of cast-in-place piles in permafrost regions is commonly estimated using the initial moisture content and frozen-soil strength parameters obtained during site investigation. However, concrete hydration heat disturbs the temperature field of the surrounding frozen soil. During thawing and subsequent refreezing, this disturbance induces unfrozen-water migration and moisture redistribution. The resulting changes in the frozen pile–soil interface may cause the measured shaft resistance to deviate from the initial design estimate. In this study, laboratory direct shear tests and engineering-oriented reduced-scale pile–soil segment tests were conducted. The effects of initial moisture content and soil stratification on interface shear strength, the surrounding temperature field, and the post-test moisture distribution were investigated. Vertical pile compression tests were also performed to evaluate changes in pile shaft resistance. The main results are as follows. (1) The shear strength of the concrete–frozen-soil interface varied nonlinearly with moisture content. It increased initially and then decreased, reaching its maximum at a moisture content of 30%. (2) The temperature rise in the surrounding frozen soil was jointly controlled by soil stratification and initial moisture content. Higher moisture contents produced smaller peak temperature rises. In the near-pile region, the maximum difference in peak temperature among soil layers with different moisture contents was approximately 10.8%. (3) During thawing and refreezing, moisture migration was jointly affected by the temperature gradient and the moisture conditions of different soil layers. Unfrozen water migrated toward colder regions or lower-moisture soil layers under temperature gradients, capillary effects, and freezing suction, resulting in near-pile moisture depletion and localized moisture enrichment. For the Group A model, the initial-state estimate underestimated the measured peak shaft resistance by 12.45%. In contrast, for the layered B1 model, the initial-state estimate overestimated the measured peak shaft resistance by 20.58%. (4) A preliminary lumped equivalent coefficient, keq, was introduced to establish a relationship between moisture content and local equivalent interface resistance. After the measured post-test near-pile moisture distributions were incorporated into the calculation, the relative deviation decreased from 12.45% to 7.36% for Group A and from 20.58% to 9.66% for B1.
Asphalt bridge deck pavements are highly susceptible to rutting, shoving, and interlayer slippage under high-temperature traffic conditions, where interlayer shear stress plays a decisive role. To clarify the coupled effects of thermal gradients and moving loads, this study developed a sequential three-dimensional thermo-mechanical finite element model for a double-layer pavement in Zhongshan, China. Field-recorded air temperature, solar radiation, sunshine duration, and wind speed were used to define transient thermal boundaries. The calculated temperature field was then transferred to a fully bonded moving-load model with dual rectangular contact areas and braking-induced longitudinal traction. Axle load, roadway slope, braking coefficient, and the thicknesses of the SMA-13 and AC-20 layers were varied. The predicted temperature fluctuation attenuated and the peak time was delayed with depth. The pavement surface reached 58.95 °C at 13:00, whereas the bottom of the asphalt overlay reached 46.99 °C at 17:00. Under the adopted 14:00 near-peak summer condition, increasing axle load amplified the overall response and raised the maximum asphalt-layer shear response from 0.172 to 0.223 Mpa. Roadway slope mainly affected traffic-direction stress transfer. Increasing the braking coefficient from 0 to 0.7 increased longitudinal shear response from 57.9 to 161.2 kPa in the asphalt layers and from 56.4 to 112.6 kPa near the AC-20/concrete interface. Increasing SMA-13 thickness reduced thermal and mechanical demand in the underlying layers, whereas increasing AC-20 thickness reduced the response near the concrete deck but shifted part of the tensile and shear demand toward the upper asphalt layer.
Headed stud connectors are critical components for ensuring composite action in steel–concrete composite structures. In practical infrastructure applications, welded headed studs may be subjected not only to interface shear forces but also to additional compressive actions induced by structural restraint, self-weight, traffic effects, and other service conditions. However, the influence of axial pressure applied along the stud height direction on the post-fatigue performance of stud-connected interfaces remains insufficiently understood. This study experimentally investigates the post-fatigue shear performance of welded headed stud connections subjected to different levels of axial pressure along the stud height direction. Nine push-out specimens were divided into three groups with axial pressures of 0, 20, and 40 kN applied to each specimen. A fatigue–static loading procedure was adopted, including initial static loading, intermittent static tests after every 400,000 fatigue cycles, and final static failure tests after two million fatigue cycles under a fatigue load range of 70–130 kN. The failure mode, load–slip response, total shear-transfer capacity, shear stiffness, and slip capacity were analyzed. The results show that all specimens maintained load-carrying capacity after two million fatigue cycles under the investigated loading conditions. The final failures were mainly characterized by shear failure at the root of the welded headed studs accompanied by local concrete crushing. Axial pressure improved the total shear-transfer capacity and shear stiffness of the stud-connected interfaces, while the ultimate slip capacity remained relatively stable. The enhancement is considered to be associated with improved interface interaction and local restraint effects, although these mechanisms cannot be independently quantified using the current test setup. These findings provide experimental evidence for evaluating welded headed stud connections subjected to combined axial pressure and fatigue loading within the investigated parameter range.
To address the technical challenges of hydration retardation and low early strength of foamed concrete in low-temperature environments of cold regions, this study investigated the effects of low-temperature curing (cycling between −5 °C and 5 °C) on the mechanical properties and microstructure of foamed concrete. Single-factor experiments were conducted to explore the effects of triethanolamine (TEA), urea, and polypropylene fibers (PPF) on the mechanical performance of foamed concrete. A response surface methodology (RSM) was employed to establish regression models between the dosages of each component and the compressive strength (CS), thereby determining the optimal mix proportion under low-temperature curing. The experimental results indicate that increasing the urea dosage leads to an increase in the flowability of foamed concrete, and the effects of the three types of admixtures on the CS all exhibit a non-linear characteristic that first increases and then decreases. The significance of the three factors on the CS of the material follows the order: TEA > PPF > urea. The obtained optimal mix proportion is 0.052% TEA, 1.08% urea, and 0.194% PPF, yielding 3, 7, and 28 d CS of 1.088 MPa, 1.342 MPa, and 2.301 MPa, respectively. Microstructural analysis via SEM observations and XRD analysis suggest that the admixtures effectively compensate for the hydration retardation induced by low temperatures, promoting the abundant generation of needle-like ettringite (AFt) and C-S-H gels that interweave into a dense network, thereby achieving higher strength. This study provides a theoretical basis and technical support for the low-temperature construction of foamed concrete subgrades in cold regions.
Pavement distresses play a pivotal role in evaluating roadway conditions and guiding maintenance strategies. In many instances, these distresses arise from construction deficiencies, substandard material quality, or inadequate maintenance practices, rather than from inherent design shortcomings. Understanding the interrelationships among different types of pavement distress is therefore essential for engineers and decision-makers seeking to enhance pavement performance and prolong service life. This study examines the statistical associations among various asphalt pavement distresses using association rule mining techniques. The dataset was collected from 18 regions in Lebanon, covering a total roadway length of 419.87 km, thereby enabling a comprehensive analysis. Ten primary categories of pavement distress were identified and analyzed using support, confidence, and Lift measures to quantify their co-occurrence patterns and dependency relationships. The results revealed strong statistical associations among most pavement distress types, with particularly strong interactions involving raveling and weathering, longitudinal cracking, alligator cracking, patching, and potholes. Raveling and weathering were the most prevalent distress factors, accounting for 20.88% of total occurrences, whereas block cracking was the least frequent, representing only 0.34%. The joint probability of raveling and weathering occurring with alligator cracking reached 37.38%, while the conditional probability of alligator cracking given the presence of raveling and weathering was 47.65%. Lift analysis further distinguished associations that were stronger than expected based on distress prevalence alone, thereby reducing the influence of frequency-driven relationships. These findings demonstrate the effectiveness of the proposed analytical framework in identifying statistically grounded pavement distress interactions and provide complementary information to support pavement management, maintenance prioritization, and resource allocation.
The post-earthquake damage state of reinforced concrete (RC) double-column piers directly affects bridge traffic capacity and emergency response efficiency. To improve the interpretability of damage assessment, this study proposes a Structural–Visual Integrated Evaluation (SVIE) framework that combines structural response analysis with image-based damage evidence. Structural responses from quasi-static tests are used to define four baseline damage states: intact-to-slight, moderate, severe, and critical damage. An improved DeepLabv3+ model is then applied to 315 global-scene images for end-to-end semantic segmentation of background, concrete spalling, and reinforcement exposure. The extracted visual evidence is used to verify its consistency with the baseline structural states. On the test set, the model effectively identified concrete spalling regions, achieving an IoU, F1-score, Precision, and Recall of 78.86%, 88.18%, 89.93%, and 86.49%, respectively. For reinforcement exposure, although IoU and Recall were relatively low because of sample scarcity and small-target characteristics, Precision reached 70.40%, indicating that detected regions can provide supplementary evidence for severe local damage. The consistency analysis showed that the morphology of visual damage was generally compatible with the progression of structural damage states. The results provide a laboratory-based proof of concept for a mechanically grounded and visually interpretable framework for rapid post-earthquake assessment of RC double-column piers.
