ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems, Part A: Civil Engineering

Crack analysis based on computer vision has become a common approach for crack detection and localization in civil infrastructure. In practice, many cracks show poor continuity, uneven gray levels, low contrast, complex topology, and background noise. These characteristics present significant difficulties for image-based crack detection. In this paper, we propose a novel framework that includes a deep fully convolutional network and densely connected conditional random field (dense CRF) to realize pixel-level crack detection in an end-to-end manner. The network learns and aggregates multilevel features at hierarchical convolutional stages. Specifically, the backbone of our network is novel self-attention modules with $1\times 1$ convolution kernels for context information extraction across channels, and the network end with multiple parallel atrous convolution filters with different rate to capture objects and features at multiple scales. Finally, we combine the network output with a dense CRF to refine the final prediction results. The network in our study is trained and evaluated using three classical crack data sets. The experimental results clearly demonstrate that our method outperforms other approaches in terms of performance.

This paper aims to study the causal relationships among several prevailing factors and the three main dimensions of the consequences of industrial accidents that involve hazardous materials. Relying on data of the Major Accident Reporting System for the period between 1979 and 2019, regression, matching and inverse probability weighting based techniques are employed to quantify the average treatment effect of many scenarios by using the standardized mean difference to assess the covariate balance. The obtained results highlight the important influence of explosions and contractors’ involvement on human casualties, major releases of hazardous materials on environmental damages, and other causes and explosions on material losses. Although human and organizational causes are involved in the three dimensions of consequences, the contribution of plant/equipment causes seems more important in the environmental and material contexts. Such results may provide a better general understanding of the causal structure in such a critical field or at least help initiate a different vision for taking advantage of the lessons learned from the occurrence of major industrial accidents.

Maritime accidents have become a major threat to societal safety and environmental protection, especially in complex navigable waters with high traffic density and diverse ship behaviors. To achieve effective safety control and efficient traffic management, a comprehensive understanding of ship behavior is essential. This study proposed a framework for ship behavior pattern analysis based on multiship encounter detection. The overall methodology incorporates research steps of data preprocessing, multiship encounter detection, and ship behavior pattern analysis. Using automatic identification system (AIS) data, the multiship encounter situations were identified and extracted. Based on the extracted encounter scenarios, the ship behavior patterns were analyzed using characteristic parameter statistics, spatial-temporal distribution mining, and spatial correlation analysis models. A case study is conducted using the historical AIS data in Ningbo-Zhoushan Port. The experiment results show that ship behavior patterns differ among the extracted encounter categories, and significant hotspots in spatial-temporal distribution can be observed. The findings on ship behaviors and traffic characteristics in complex navigable waters provide theoretical references for maritime traffic management authorities to mitigate risks and improve maritime safety.

A novel framework was developed for the stability analysis of spatially variable slopes under rainfall by approximating the performance function in standard normal space by a second-order surface at the design point. Saturated hydraulic conductivity ($k_s$) and normalized rainfall intensity ($Q$) were considered in the shear strength equation. The effects of drying and wetting soil water characteristic curves (SWCCs) were also considered. The most conservative model was proposed out of the five hydraulic conductivity models considered in the analysis. The concepts of spatial variability were then applied to the second-order reliability method (SORM). The random fields, gamma for stability number ($c^{\prime }/\gamma H$), lognormal for effective friction angle (${\varphi }^{\prime }$), nondimensional inverse of air entry values ($\alpha \gamma H$), and slope of SWCC ($n$), and inverse Gaussian rainfall intensity ($Q$) were generated for clayey slopes. The results revealed that the percentages difference ($\epsilon$) in reliability index (${\beta }_{\mathrm{SORM}}$) obtained using drying and wetting SWCCs are 19.39% and 37.17% for $Q=0$ and 0.5, respectively. The findings also illustrated a considerable shift in the critical slip surfaces (CSSs) due to the change in rainfall intensity. Practically, the vertical fluctuation scale was estimated more precisely than the horizontal one. The results show that the influence of vertical autocorrelation distance (ACD) on the reliability index is more significant when compared with horizontal ACD. However, it is noted that the influence of vertical ACD (${\delta }_y$) greater than 3.0 times the height of the slope on ${\beta }_{\mathrm{SORM}}$ stabilizes. Also, the spatial variations in soil properties along the horizontal direction are relatively more uncertain due to the heterogeneity of soil deposits, sampling limitations, and difficulty of characterizing large-scale spatial variations. This uncertainty in the horizontal variability should be given due consideration in addition to vertical variability because it may have a greater impact on the estimation of the reliability index and overall behavior of the slopes. A parametric study was conducted to estimate the anticipated performance levels of rainfall-triggered slope failures. Additionally, a case study of slope failure triggered by rainfall in the Konkan region of Maharashtra in India is presented at the end. It is noted that the Konkan slope may not be stable when $Q=0.83$.

This paper aims to use blockchain technology to promote the development of the “single contract” system to solve the critical problems of multimodal transport process optimization. Firstly, we study the development process and current research status of blockchain technology and multimodal transport “single contract” systems. The bottleneck of developing a multimodal transport “single contract” system in the blockchain application in the sharing platform is identified. Based on the aforementioned research, and combined with Petri network theory, we propose to establish a multimodal transport process model and optimization model by taking sea-rail intermodal transport as an example. It also presents the construction of a blockchain-based information-sharing platform for a multimodal transport “single contract” system. This paper applies blockchain technology to multimodal transport business, enabling different enterprises and departments to integrate their business, realize the barrier-free transmission of information, and shorten the business process time, thus promoting the construction and development of multimodal transport system.

This paper presents a novel method for obtaining analytical solutions for the nonstationary response of multidegree-of-freedom (MDOF) systems endowed with half fractional derivative elements and subjected to external stochastic excitation. Specifically, first, the proposed technique employs eigenvector expansion of the state-space formulation and Laplace transforms to derive an analytical solution for the impulse/frequency response function (IRF/FRF) of the fractional-order dynamic system. Moreover, by utilizing the Laplace transform method, exact analytical solutions for the second-moment response are obtained in the frequency domain. A comprehensive set of six numerical cases is presented to demonstrate the effectiveness of this novel methodology. These cases include two degenerated scenarios, namely a single-degree-of-freedom (SDOF) system and a two-degree-of-freedom (2-DOF) linear system, both endowed with half-order fractional derivative elements and subjected to stochastic stationary/nonstationary excitations, including white noise, modulated white noise, and modulated colored noise with modified Kanai-Tajimi spectrum. The analytical nonstationary responses derived by the proposed method exhibit exceptional agreement with pertinent Monte Carlo simulations, validating the accuracy and reliability of the proposed approach.

Cooperative obstacle avoidance of multiple unmanned surface vessels (multi-USV) is an important issue in unmanned vessel safety. This study aims to solve dynamic safety collision avoidance problems, especially decision oscillations, which exist in the traditional velocity obstacle method for multi-USV cooperative obstacle avoidance. To cope with these problems, we analyze the principle of the reciprocal velocity obstacle method based on multiple constraints, such as the dynamics constraint and International Regulations for Preventing Collisions at Sea (COLREGS) constraint of unmanned vessels, and propose an improved velocity obstacle method under multiple constraints. The method introduces the uncertainty factor of the velocity obstacle to obtain the optimal velocity of the USV when performing obstacle avoidance actions on the water surface and then achieves the replanning of the trajectory. Through simulation experimental verification, we demonstrate that this method can effectively resolve different encounter conflict situations, achieve coordinated motion control among multiple USVs, and improve the safety of USV obstacle avoidance. In addition, the method meets the dynamic obstacle avoidance requirements of surface unmanned vessels under multiple constraints in actual navigation. This research is of great significance in improving unmanned vessels’ autonomous obstacle avoidance capability and provides strong support for the practical application of unmanned vessels. Future research can further explore the applicability of the method.

Destructive earthquakes are rare, high-impact, uncertain events. When they occur, the physical infrastructure is damaged, resulting in major economic losses. Risk mitigation planning aims to prevent extreme adverse effects and address tail risk; however, it requires large up-front investments. The large underlying uncertainties generate high variation in the avoided losses. Disregarding these uncertainties does not eliminate their presence nor detract from their importance. Two distributionally robust optimization (DRO) models are proposed to select building groups for pre-earthquake retrofit considering the uncertainties in the (1) earthquake occurrence probabilities, and (2) within-scenario building damage costs. The models minimize the worst-case expected objective function cost given the uncertainty in the random variables, promoting informed decisions under incomplete information. The conditional value at risk (CVaR) measure is incorporated into the optimization framework to model the cognitive loss-averse bias in decision-making for low-probability, high-consequence events. CVaR is derived by taking a weighted average of the extreme damage costs in the tail of the distribution, beyond the value at risk cutoff point, refining previous research that measured risk by setting arbitrary thresholds that are hard to define in practice. Implemented for the city of San Francisco, the risk-based models guard against high damage costs at the right tail of the distribution at the expense of higher up-front costs. The objective function cost was evaluated using out-of-sample data to assess the model performance under unseen data. The DRO reformulations resulted in improved model performance in the out-of-sample testing relative to the nonrobust approach, mitigating the optimizer’s curse, but may lead to overly cautious retrofit decisions if uncertainties are overestimated.

The process of a concrete target subjected to a tandem warhead was studied by means of field testing and numerical simulation. First, field testing of a shaped charge impacting the concrete target was carried out, in which the concrete targets included normal strength concrete and ultrahigh performance concrete. Then, the rigid projectile was used to penetrate the damaged concrete targets experimentally. Numerical simulation was adopted to simulate the two stages of a tandem warhead penetrating into the concrete targets. The combined effect of penetration and explosion in the open literature was also reproduced by simulation. By comparing the depth of test data and simulation results, the accuracy of the algorithm, material models, and corresponding parameters was proven. Finally, the numerical simulation method was adopted to simulate the process of a tandem warhead damaging a concrete target, including shaped charge impaction, penetration following the rigid projectile, and explosion inside the concrete. As seen from the comparative results, the parameters used in numerical simulation can appropriately describe the concrete damage, and the tandem warhead can effectively improve the damage degree of concrete targets. The greatest influence of concrete strength on the damage subjected to a tandem warhead is the stage following projectile penetration. The conclusion of this paper can provide important reference significance for the design of tandem warheads and protective engineering.

In recent years, liquefied natural gas (LNG) gradually has become an alternative fuel for ships. For targeted safety management of ship-to-ship LNG bunkering, this study developed a new method to identify, quantify, and rank the risk influential factors (RIFs) for fuel spills during the process of ship-to-ship LNG bunkering. Firstly, starting from the process of ship-to-ship LNG bunkering, the fuel leakage RIFs of ship-to-ship LNG bunkering were identified and summarized. Secondly, combining failure mode and effect analysis (FMEA), Bayesian networks (BN) based on the fuzzy belief rule (FBRBN), and evidential reasoning (ER), a risk assessment model is proposed to quantify the risk level of the RIFs. Finally, using the case study of Zhoushan LNG bunkering station, China, the feasibility and practicability of the established risk evaluation index system and research methods were verified. The results of this study show that improper handling by personnel is the most important RIF affecting the safety of ship-to-ship LNG bunkering. Based on the results, targeted preventive measures are proposed to enhance the safety of ship-to-ship LNG bunkering.

A mitigation system is significant in that it blocks the flow of toxic gas and prevents the growth of hazardous areas, thereby securing sufficient evacuation time. A new air curtain mitigation system was developed to prevent the dispersion of toxic gas by breaking the momentum of flow and reducing the extent of the affected region. To investigate the effect of key factors, computational fluid dynamics (CFD) simulations were performed for various values of air curtain angle and velocity. The air curtain velocity was adjusted at 0, 30, 40, and $50\mathrm{m}/\mathrm{s}$, whereas the air curtain angle was varied at 0°, 10°, 20°, and 30°. In this study, the mechanism of the air curtain by which the linear motion of the released toxic cloud is converted into rotational motion was clarified. The transformed rotational motion was evaluated by vorticity, and its value increased as the air curtain velocity increased. Increasing the air curtain velocity improved the mitigation effect at a fixed air curtain angle. Increasing the air curtain angle contributed to a dilution of toxic gas concentration by enhancing the mixing with air and breaking the momentum of flow. Based on the simulation results, the toxic gas mitigation effect of the air curtain system is confirmed, and air curtain design guidelines are presented.

Given the impact of climate change on the society, climate adaptation for civil structures and infrastructure systems has become an urgent task. To determine an adaptation and maintenance strategy in a rigorous manner, the influence of climate change on structures and infrastructure systems should be quantified. Risk, a performance indicator incorporating both the failure probability of structures and their associated failure consequences, is suitable to be used to characterize the effect of climate change on the performance of structures and infrastructure systems. To date, several studies have been published on climate change effects on structures as well as the associated adaptation and maintenance strategies. However, most of those studies focused on extreme events such as hurricanes and floods, whereas there is little research on the influence of climate change on progressive deterioration mechanisms such as corrosion. In addition, these research efforts placed a greater emphasis on the performance of individual structures than on the performance of groups or networks of structures. In this paper, the effect of climate change on a bridge network (an important type of infrastructure system) subjected to corrosion is quantified. Girder replacement actions are adopted as the maintenance policy. Optimal maintenance solutions for the bridge network considering the effect of climate change were determined using optimization algorithms. The results show that the climate change can have a significant influence on the network risk and can alter the optimum maintenance schedule under a specific maintenance budget.

The prewarning and responses of different monitoring indices are out of sync in engineering disaster warning, and the disaster risk assessment is inaccurate based on individual response index or comparison with different indices. The traditional Dempster–Shafer (DS) evidence theory cannot readily integrate the conflicting multivariate monitoring data. In the present study, the DS evidence theory was improved by integrating various conflicting multivariate monitoring data, and the application condition, advantages, and disadvantages of those modified methods based on the DS evidence theory were investigated. An improved DS evidence theory method was proposed based on the Chebyshev distance and the zero-divisor modified evidence source method. The results indicated that the improved DS evidence theory based on the Chebyshev distance performs well in both integrating the conflicting and nonconflicting monitoring data and is superior to other improved methods in suppressing interfering evidence with good stability. The proposed improved DS evidence theory based on the Chebyshev distance is then applied to rock burst prewarning, and the prewarning model is established based on multiphysics in situ monitoring data. The probability with various risk levels is employed to assess the safety state, which can reflect the degree of rock burst. The risk of rock burst can be quantitatively predicted using this proposed method, which can provide some guidance in the prewarning of engineering disasters.

Model factors for barrette piles at the ultimate limit state under compression loading were evaluated in this study. A total of 64 rectangular pile load test results from different locations around the world were compiled as CYCU/Barrette/64. The database was divided into a drained group and an undrained group corresponding to the dominant soil conditions along the shaft length. The measured capacity was interpreted from a load–displacement curve using nine common methods. The model factors for barrette pile under compression loading were determined to provide engineers with the statistics to implement reliability-based design (RBD). It was found that the drained model factors ($Q_m/Q_p$), where $Q_m$ = interpreted capacity and $Q_p$ = predicted capacity, range from 0.57 to 1.09 and the undrained model factors range from 0.64 to 1.31. A model factor less than 1 is unconservative (predicted capacity is larger than the interpreted capacity). These model factors can be used to calibrate resistance factors for RBD at the ultimate limit state (ULS). Hyperbolic model factors were also determined to calibrate deformation factors for RBD at the serviceability limit state (SLS) at any tolerable displacement value. The mean hyperbolic model factors were determined as (1) mean $a=15.65\mathrm{mm}$ and mean $b=1.07$ under drained conditions; and (2) mean $a=15.86\mathrm{mm}$ and mean $b=0.81$ under undrained conditions. The load–displacement curves are fitted to these hyperbolic model factors ($a$ and $b$) using $Q/Q_p=\rho /(a+b\rho )$, where $Q$ = applied load and $\rho$ = pile head displacement. It was found that the model factors might depend on some input parameters, although there are insufficient load tests to establish this preliminary finding conclusively.

The conventional surrogate model for slope reliability assessment often is faced with the issues of high dimensionality and sample selection disorder, which are caused by the spatial variability of soil parameters and which compromise the precision and efficiency of slope reliability assessment. Previous studies focused on solving this problem mainly by choosing more-accurate models; studies of optimizing the training samples for constructing surrogate models are relatively scarce. This paper proposes a multivariate adaptive regression spline model based on active learning (AMARS) for slope reliability analysis in spatially variable soils, combined with the sliced inverse regression (SIR) method. The active learning includes self-supervised learning methods that optimize the sample set for constructing surrogate models. The training samples are processed using the SIR method to prevent the model from falling into dimensionality disaster. The proposed method was validated using two slope cases with spatial variation. Comparison of computational efficiency and accuracy in estimating slope failure probability revealed that the method suggested here outperforms others. Moreover, for both single-layer simple and multilayer complex spatially varying slopes, the proposed method not only reduces computational costs effectively but can also be used to evaluate the reliability of slopes with small failure probabilities.

When a ship accident occurs, emergency evacuation of passengers in a shorter time is one of the most effective means of reducing casualties. However, in addition to the ship inclination, the efficiency of the emergency evacuation can be affected by human behavior (e.g., competitive behavior), which affects human moving speed. Therefore, to analyze the impact of competitive behavior during ship evacuations, a dynamic evaluation system is developed to measure nested competitive behavior. Firstly, the perceived area is obtained by dividing the pedestrian visual perspective, which is used to calculate crowd density at every time step. Secondly, fuzzy logic is used to calculate the real-time competitive degree based on the inclined angle and crowd density, integrated into the human evacuation model as an input parameter to update competitive behavior. Finally, this study analyzes and evaluates evacuation time and efficiency with different proportions of competitive people at different inclined angles, using a dining room on a ship as a case study. The results show that without ship inclination, the total evacuation time decreases with an increase of the proportion of competitive people as more competitive people can accelerate the evacuation process. However, the inclination of a ship leads to a decrease in human walking speed, congestion at the exit, and a slower overall evacuation process. According to the findings of this study, an appropriate increase in the proportion of competitive humans is beneficial to the efficiency of emergency evacuation, while strengthening the guidance at the exit will also reduce the evacuation time.

Projections using global climate models indicate that climate change will influence the patterns of natural hazards, such as thunderstorms, atmospheric river landfalls, extreme droughts, and ocean waves. The frequency and intensity of these hazards are expected to increase gradually in proportion to global temperature. The design principles based on the philosophy of cost optimization need to be updated to accommodate the nonstationarity of the load processes, primarily because the prevalent cost analysis methods in the literature predominantly assume that the loads are stationary. This study provides a novel methodology for calculating the first two moments and the distribution of the economic losses for nonstationary loading processes. Here, the load processes are modeled as a nonhomogeneous Poisson process (NHPP) with time-dependent rates. The presented methodology is applied to estimate the losses due to tornadoes in Ontario, Canada and heat waves in US cities. It was found that if adaptive measures are applied to increase the capacity of structures, the losses due to these climate-driven hazards can be significantly reduced. For example, if mitigation strategies are employed in Ontario, such that the effect of tornadoes with wind speeds lower than $50.3\mathrm{m}/\mathrm{s}$ becomes negligible, then the tornado losses until 2100 can be reduced by 66%.

Abstract

In the laying of long-distance pipelines, it is sometimes impossible to avoid one or more areas that are prone to frequent geological disasters, such as landslides. In the case of such a disaster, the buried pipeline is likely to undergo large displacement leading to plastic deformation, subsequent leakage, explosions, and other accidents that may result in its failure. In order to ensure the safety of pipeline transportation, in this work, a remote real-time system for monitoring the status of pipelines was designed on a cloud service platform to realize stress–strain analysis and to provide an early warning of pipeline damage after a landslide. The results of the stress–strain analysis of a pipeline buried under a landslide were used to establish a numerical calculation model based on the shell element and nonlinear soil springs. The deformation distribution characteristics of the pipeline, based on multiple factors, were studied, and the effects of the landslide width, buried depth, ultimate soil resistance, diameter thickness ratio, and internal pressure on vertical displacement, as well as the axial strain and bending strain of the pipeline were obtained. According to the results of the finite-element method, the plastic deformation position of the pipeline under the action of landslide was determined, the software and hardware configuration of the pipeline strain monitoring scheme was designed, and the installation of the pipeline strain monitoring system was carried out. The processing results of the field data showed that the model had a good noise reduction effect. Moreover, the results showed that the system achieved stable real-time data acquisition, efficient data remote transmission, convenient operation, and rich terminal monitoring capabilities, thus effectively providing an evaluation of the operating status of the pipeline, improving landslide disaster warning, and ensuring the safe operation of the pipeline.

Practical Applications

Buried pipelines are likely to undergo large displacement under the action of geological disasters such as landslides, which can lead to accidents such as pipeline leakage and explosion. In order to ensure the safety of pipeline transportation, a numerical calculation model of a buried pipeline based on shell elements and soil springs is established to analyze the stress and strain of a pipeline under a landslide. The model can reflect the deformation distribution characteristics of the pipeline and analyze the influence of landslide width, buried depth, and other factors on the deformation of the pipeline. Based on the presented method, the dangerous points of plastic deformation of pipeline under landslide can be determined. Furthermore, combined with the actual situation of a landslide site, a monitoring system has been designed and installed that can operate stably for a long time in the landslide disaster site. The system can realize the acquisition, transmission, and evaluation of pipeline data, and ensure the smooth operation of buried pipelines.

Distributed energy resource systems (DERs), such as rooftop solar panels, are gaining traction in the energy sector to improve infrastructure resilience to earthquakes by facilitating electricity sharing in seismic regions (e.g., California). The seismic risk assessment of DERs requires estimating ground motion intensity measures (IMs) using ground motion models (GMMs). In this context, a few existing efforts have used ergodic GMMs, which assume that the distribution of IMs over time (across multiple events) at a single site is the same as that of ground motion IMs over space (across multiple sites). However, with the advent of large ground motion databases, it has become evident that ground motions are influenced by location-specific repeatable effects, motivating a gradual transition into nonergodic approaches that can capture these effects. However, the impact of these approaches (i.e., ergodic and nonergodic) on the seismic risk assessment of DERs has not been assessed. This study considered areas with contrasting spatial extents to assess the impact of nonergodic approaches in the seismic risk assessments of DERs. Specifically, we investigated the risk of power outages in residential communities that have access to DERs and are exposed to a significant seismic hazard. The results indicate significant differences in the estimated risk (as much as 0.2 on a scale of 0–1) between ergodic and nonergodic estimates at locations where large intensity measures and significant repeatable effects are observed. Furthermore, the nonergodic approach is better equipped to capture the spatial variation of risk estimates across a large spatial extent, but more data are required to fully realize the potential of nonergodic approaches in community-scale regions.

Coastal structures such as seawalls prevent hinterland erosion and flooding, making them an integral component in building resilient communities and infrastructure. However, these structures could fail to perform as intended during extreme events, given that their performance varies with environmental conditions such as incident wave characteristics. The probability of failure may be reduced using the flood mitigation benefits of natural or nature-based features, such as beaches. However, the performance of hybrid solutions is not yet fully understood. The objective of this study is to quantify the wave attenuation benefits of beaches for improving the performance of seawalls to reduce wave runup and prevent wave overtopping. To that end, this study first proposes novel metrics for the robustness and serviceability of such systems, as well as a new temporal metric to quantify the flooding time. Next, these metrics, as well as the volume of overtopping, are utilized to measure the performance of a hybrid beach–seawall system at a study site in the United States. A numerical modeling approach, using the phase-resolving wave model XBeach nonhydrostatic, is adopted to generate a comprehensive dataset of wave runup and overtopping. The results of numerical experiments are analyzed to investigate the value of the beach in reducing wave runup and overtopping on the seawall and, thus, enhance the robustness and flood serviceability of the structure during extreme events. The analyses are carried out for various scenarios of storm tides and wave characteristics and beaches of different widths and slopes. The results reveal that the influences of beach characteristics are significant in a hybrid beach–seawall coastal defense system, with a gentler (wider) slope reducing wave runup and overtopping, particularly for smaller storms and low storm tides. However, a steep (narrow) beach offered greater robustness against larger storms and higher storm tides, and the system’s robustness and serviceability were highly sensitive to the storm tide across all beach slopes, implying a nuanced approach to the design and management of the systems.

Identifying the mechanical properties of civil structures is required for life-cycle assessment. Kalman filters are exploited for this goal, enabling the online update of a numerical model, acting as the digital twin of the structure, and quantifying the uncertainty of the estimated properties. As uncertainty about model formulation is usually disregarded in the identification, model class evidence has been recently formulated to compare different parametrizations of the properties of the monitored structure through a metric, allowing selection of the most plausible one. When dealing with parameter estimation, predominantly model evidence is deployed in batch Bayesian estimation. Here, the formulation of model class evidence is proposed for the unscented Kalman filter, which allows online calculation of model class evidence for a system without the need to compute the mapping gradient in time. This formulation was inspired by the model class evidence developed for the extended Kalman filter. Numerical results related to shear buildings are presented to validate the metric, showing the impact of under- and over-parametrizations on identification.

Corrosion is a major threat to RC bridges for their lifelong safety and sustainability. This article investigates the spatiotemporal progression of corrosion degradation in piers of a RC bridge and estimates its impact on life-cycle seismic resilience of the bridge. Spatial randomness of corrosion and concrete parameters along the length and periphery of piers were generated through the spectral representation method. Generated random fields were integrated with the multiphysics phenomena of corrosion in the COMSOL Multiphysics platform for a realistic simulation of corrosion propagation on rebars in bridge piers. Obtained information was then inserted to the finite-element model of the bridge composed of all essential features to accurately capture this gradual degradation for seismic performance assessment of the bridge in a life-cycle context. In this performance simulation, the study considered both spatial and nonspatial corrosion strategies in the presence and absence of early-age surface cracks and estimated the variation in measured seismic resilience of the bridge for various corrosion strategies. Results depict significant overestimation of seismic resilience of the bridge along its life span if the spatial nature of corrosion is ignored, and thus highlights the importance of considering the same for seismic response analysis of aging bridges.

Abstract

Manifold calamitous consequences of seaport hazardous cargo accidents have raised grave trepidation over the safe management of hazardous cargo. Fewer investigations have aimed at the multifaceted risk associated with intricate seaport hazardous cargo loading and unloading operations. To that end, this study analyzes the multifaceted risks associated with seaport hazardous cargo loading and unloading operations. Bayesian network (BN) is employed and amalgamated with interval type-2 fuzzy sets to account for intricacies and uncertainties concomitant to the causation factors. Power average operator on aggregation of domain expert opinions, and experts scores in quantifying conditional probability tables are employed to further augment the viability and diagnosis accuracy of the analysis. The results reveal policy and regulations failures, errors and violations, and communications issues as the prominent accident causation factors. To identify the critical accident causation factors, a sensitivity analysis was performed. Based on these insights, the study proposes practical applications in terms of development and abidance of rules, resource upkeeping, and critical operations to mitigate the risk of seaport hazardous cargo loading and unloading operations.

Practical Applications

The study’s outcomes propose practical measures to mitigate risks effectively. Adhering to safety regulations is paramount, especially when handling petrochemicals and ammonium nitrate. Staff must diligently follow rules to prevent negligence or authority misuse. Monitoring personnel qualifications and experienced supervisors play a vital role in promptly identifying threats. Consistent checks and maintenance of technology, infrastructure, and equipment are indispensable. Ship positioning entails assessing water depth and vessel draught. Secure mooring and monitoring movements due to tides, currents, and operations are vital. Seamless coordination between ship and terminal via designated communication systems is central, including comprehension of emergency signals. Cargo information—grade, quantity, contaminants, and local regulations—is crucial. Uncontaminated cargo holds are vital. Likewise, creating a pre-arrival cargo plan, approved by the terminal and shared with ship and terminal teams, is highly critical. Deviations require authorization, with copies sent to the port agency. The plan must detail loader capabilities and rates. Ballast efficacy, frozen ballast readiness, and equipment accuracy need rigorous checks. Adjustments are needed if the ship’s cargo acceptance rate differs from the terminal’s loading capability.

Abstract

As a promising clean fuel, the liquefied natural gas (LNG) has been considered as the alternative energy for ships and the LNG-fueled vessels has sharply increased in recent years. However, the leakage of LNG has become a serious challenge for the wide application of LNG-fueled vessels, especially during ship-to-ship (STS) bunkering. This study analyzes leakage accidents during STS LNG bunkering, considering its operation process, and develops a fuzzy Bayesian network model to evaluate the probability of LNG leakage during STS LNG bunkering. Sensitivity analysis is carried out to identify the critical risk factors. Moreover, the study simulates the consequence of LNG leakage using the Fire Dynamics Simulator (FDS) software and derives the F-N curve to determine the risk levels of LNG leakage considering the thermal radiation criteria. A case study of LNG-fueled vessel in Ningbo is conducted to verify the proposed method. The results show that the probability of LNG leakage and fire accidents during STS bunkering is 0.93% and $1.3829\times {10}^{-4}$ respectively, which is located in the as low as reasonably practicable (ALARP) zone. Afterwards, it is found that uneven force on the ropes and fenders are the crucial factor influencing LNG bunkering. The simulation experiments demonstrate that the high-risk areas of STS LNG bunkering are 4 to 16 m away from the LNG-fueled ship.

Practical Applications

This study analyzes leakage accidents during ship-to-ship (STS) liquefied natural gas (LNG) bunkering, and develops a fuzzy Bayesian network model to evaluate the probability of LNG leakage during STS LNG bunkering. A case study of LNG-fueled vessel named XINAO in Ningbo is conducted to verify the proposed method. Moreover, the study simulates the consequence of LNG leakage using the Fire Dynamics Simulator (FDS) software and derives the frequency versus number of fatalities (F-N) curve to determine the risk levels of LNG leakage considering the thermal radiation criteria. The results show that the probability of LNG leakage and fire accidents during STS bunkering is 0.93% and $1.3829\times {10}^{-4}$ respectively, which is located in the as low as reasonably practical (ALARP) zone. Then, it is found that uneven force on the ropes and fenders are the crucial factor during STS LNG bunkering. The simulation experiments demonstrate that the high-risk areas of STS LNG bunkering are 4–16 m away from the LNG-fueled ship. This paper delivers a remarkable research work providing rule-makers with an insight into the safety management during STS LNG bunkering.

Mixed traffic often results in large deviations in bus schedules due to the influence of stochastic factors on bus operations. In order to improve bus schedule adherence, signal priority is often employed at signalized intersections by adjusting signal timing. However, traditional deterministic optimization methods that treat bus travel time as a constant are not effective in meeting the challenges posed by changes in bus operations. To address these limitations, this paper proposes a robust signal priority control method that aims to improve the reliability of bus operations. However, because of the limited adjustable range of signal timing, signal control used independently only works a little in large bus headways. In order to further enhance control performance, bus holding is introduced in this study, leading to the development of an integrated robust signal priority and holding control (RHC) method. The RHC method improves bus schedule adherence and reliability in bus operations and is designed based on a bus schedule deviation model that takes stochastic bus speed and the number of queuing vehicles at intersections into account under mixed traffic. Through simulation experiments, the effectiveness of the proposed calculation models is verified. Furthermore, the RHC method can boost bus schedule adherence by 38% while limiting the negative impact on car delays to only 3%.

The study of the statistical law of the response spectrum of bridge structures under the action of random traffic flow loads can not only evaluate the service performance of bridge structures, but also verify the rationality of the design load taking and provide a reference basis for the control of bridge traffic flow. In this paper, the moving vehicle loads are simplified to moving constant forces, and the statistical characteristics of displacement response spectra and acceleration response spectra of an assembled hollow-slab girder bridge under the action of statistical steady-state traffic flow loads, as well as their spatial distribution laws over the whole bridge, are investigated based on the existing statistical distribution model of traffic flow parameters, which are obtained by Monte Carlo simulation for random traffic flow loads. The results show that with the increase of statistical samples, the mean, standard deviation, and probability density of the steady-state response spectrum estimator of each monitoring point gradually converge, and its convergence value is only related to the random traffic flow load model and the bridge structure characteristics; the convergence values of the mean and standard deviation of the response spectra are parabolically distributed along the longitudinal direction of the bridge; the response spectrum estimator of different precast beams at the same cross section are different due to the different loads transferred laterally. The statistical properties of the response spectrum estimates are not affected by accidental external factors, and can be used as indicators to determine the operating conditions of the bridge structure and the magnitude of the traffic flow on the bridge deck, providing a reference basis for the health monitoring of the bridge structure.

The spatial variability of soil properties is pervasive, and can affect the propagation of seismic waves and the dynamic responses of soil–structure interaction (SSI) systems. This uncertainty is likely to increase the damage state of a structure and its risk of collapse. Additionally, conducting multiscale simulations efficiently in the presence of uncertainties is a pressing concern that must be addressed. In this work, a 3D probabilistic analysis framework for an SSI system considering site effects and spatial variability of soil property (i.e., elastic modulus, $E$) has been proposed. This framework is based on the random finite element method (RFEM) and domain reduction method (DRM). A multiscale model of a five-story reinforced concrete (RC) frame structure was developed on an ideal 3D slope to verify the effectiveness of the proposed framework. The dynamic responses of the structure were analyzed, and the peak floor acceleration (PFA) and peak interstory drift ratio (PSDR) were selected to estimate the damage state of structures. It was found that the proposed method significantly improves computational efficiency approximately 20 times compared with the direct method. In the regional models, with the increase of the coefficient of variation (COV) of $E$, the energy of seismic waves becomes more concentrated at the crest and the response spectrum value of medium and long periods increases. In the local SSI model, the soil variability increases the mean value of PSDR, resulting in a more severe damage state compared to the results from the deterministic analysis. Consequently, this study provides some suggestions for engineering practice, and the importance of probabilistic analysis considering spatially variable soils in the SSI problem is highlighted.

Crack fracture in large rotating shaft results in catastrophic failure in many industrial plants. To prevent this, a health index model is developed to diagnose the crack severity by the vibration sensor. A full-scale test rig was constructed with a similar configuration to the field conditions to reduce the associated uncertainties. Vibration signals were collected under various operating conditions for artificial cracks with different severities and locations. After identifying useful features and operation parameters, which include 1st shaft harmonic frequency (1X), 9th frequency domain feature (FF9), and the rotating speed, random forest regression was employed to establish the health index model using them. Uncertainties were accounted for in determining the thresholds of normal and crack conditions. In order to apply the model in the field, features from the field equipment were scaled to those of the test rig in a statistical manner. Health ranges for the normal, caution, and danger with respect to the crack severity are recommended for practical applications.