A probabilistic structural analysis framework employing a generalised polynomial chaos expansion (gPCE) surrogate model is presented. The framework reduces computational cost and enables efficient global sensitivity analysis, Bayesian finite element (FE) model updating and uncertainty quantification (UQ). The case study under consideration is a pedestrian and cycling bridge with twin steel arches and a suspended steel-concrete deck. The ambient vibration measurements provided fifteen vibration modes whose natural frequencies were used for Bayesian FE model updating. This allowed the identification of parameters predominantly influencing higher frequencies. The model updating resulted in a reduction of uncertainties related to material parameters, deck mass and support stiffnesses. The study demonstrates the effectiveness of the gPCE surrogate model for probabilistic analysis of a real-world bridge.
Atmospheric corrosion is a critical driver of steel degradation in natural environments, and thus accurate assessment of the corrosion condition is essential for ensuring engineering safety. However, empirical models have limited applicability and poor transferability across different materials and environments. Traditional machine learning models, on the other hand, have weak expression and generalisation ability, along with poor interpretability. Therefore, an attention-augmented convolutional neural network (CNN) was developed. Firstly, a database of corrosion samples containing 24 types of steel, such as carbon, low alloy, and weathering steel, was established. Secondly, a one-dimensional CNN was formulated to extract complex high-dimensional features from sequential data, while two attention mechanism modules are incorporated to further enhance the representation of key information. Then, the interpretability of the model was analysed using the SHAP method. Finally, the long-term structural behaviour of a steel bridge was evaluated by coupling the intelligent model with numerical analysis framework. Results indicate the attention-augmented CNN model has higher accuracy than the existing models, with the evaluation metrics in test data being MAPE = 12.50%, RMSE = 5.45, MAE = 3.28, and R 2 = 0.930, respectively. This study can provide a scientific basis for the long-term performance evaluation, and maintenance planning of infrastructure engineering.
In recent years, large-scale retrofitting of ageing infrastructure reinforced with plain bars has become increasingly common in Japan. A critical challenge during this process is the temporary hookless state of plain bars. However, research data on its mechanical behaviour is limited. To address this gap, this study investigates the flexural behaviour of a hookless plain bar reinforced concrete slab by conducting two sequential four-point bending tests in orthogonal directions on the same specimen. Multi-point strain gauges and displacement transducers were used to measure the plain bar strain and relative slip between plain bars and concrete. Measurements revealed that the loss of bond strength occurred progressively, and pre-existing through-cracks led to earlier plain bar slip. The peak load was reached when relative slip initiated in all monitored plain bars. Compared to theoretical flexural capacities calculated for deformed bars, the plain bar slab exhibited experimental capacities that were 18% lower in the first bending test and 35% lower in the second. The more pronounced reduction in the second test was likely exacerbated by pre-existing cracks and other experimental factors. These findings provide crucial data for the retrofitting design of existing plain bar reinforced concrete structures.
Global climate change increases the risk of flash floods in urban areas of arid regions. Low-capacity storm drainage systems (SDS) designed for short return periods and clogged sewers during long dry periods intensify the severity. The present framework categorises the service area into different zones to assess SDS's resilience to varying return-period floods using hazard, vulnerability, and exposure indices. The hazard index evaluates the hydro-geomorphological components based on water depth in manholes as a function of service area surcharge and overflow. The vulnerability index considers land use type, and the exposure index incorporates population density. The hierarchical approach estimates the risk index for each zone and the overall flood risk of a region. Real-time control (RTC) employs evidential reasoning within control structures to minimise flood risk by leveraging the capacity of manholes. The framework's application to an 11.32 km2 urban area in the Qassim Region of Saudi Arabia demonstrates the RTC's pragmatism for 5-, 10-, 20-, 50-, 100-, 200-, and 500-year floods. The existing SDS exhibited varying levels of resilience to flooding risk, ranging from very low to very high, affirming the approach's robustness. The framework will assist municipalities with mitigating flood risk using RTC in urban areas of arid regions.
The ingress of chlorides into concrete structures can initiate corrosion, leading to reduced structural reliability and damage phenomena such as cracking and spalling of the concrete cover. This study presents a framework to identify the most cost-effective non-structural intervention from a pre-selected set of possible interventions to minimise the total maintenance cost over the intended lifetime of the structure. The maintenance cost involves the cost of the possible non-structural intervention, the expected cost of corrosion-induced spalling and/or cracking and the expected cost for a structural rehabilitation, which is imposed to be required upon exceeding a critical corrosion degree threshold. The framework explicitly models the influence of the interventions on the ingress of chlorides and integrates additional information from visual inspections and chloride profile measurements in a Bayesian framework. A case study of a reinforced concrete beam subjected to chloride ingress is considered where the decision-making framework is applied to determine the most optimal intervention. In a pre-posterior analysis, the decision-making framework is applied to investigate the influence of additional information, design parameters and the incorporated costs on the expected maintenance cost and proposed intervention strategy. The value of information of a chloride profile and visual observations is quantified.
Assessing the risks from extreme events, and deciding on the appropriate measures to ameliorate these risks, may be challenging for many policymakers who are unfamiliar with quantitative measures of risk. There is a need for risks to be expressed in terms understandable to policy makers and the public. An illustrative case study is a risk assessment of 14 fig trees from an iconic street in Newcastle where their safety risks were judged too high because an arborist consultant calculated that the risk of harm was 1 in 19.8 per tree per year - so walking down that street was 400 times more dangerous than driving a car! Other examples relate to terrorism and climate change. There is often an inability of the community and public officials to understand the notion of risk and probabilities. The paper will describe how decision metrics often involve probabilistic thinking by public officials, and how best to communicate this type of information in a manner more easily understood by the community and public officials.
Vessel collision is a critical accidental load in bridge design, while pile scour undermines structural capacity and durability. Limited studies were carried out to investigate scour progression on impact load response. To evaluate the combined action, a safety assessment for a cable-stayed bridge is conducted in this study. Finite element models of three impacted reinforced concrete columns are established and validated. A refined ship-bridge collision model incorporating dynamic pile-soil interaction is developed. The impact force, the dynamic response and the associated damage variations of the bridge tower are examined under different conditions of scour depth, water level, and collision location. The safety of the bridge is evaluated by the vessel collision safety factor determined by the actual resistances of the piers and the internal reaction. Key findings indicate that: (1) water level fluctuation has more significant influence on peak collision force than scour depth; (2) bridge displacement increases substantially with scour depth, while pile head damage exhibits a distinct non-monotonic trend; (3) the pile foundations and pile-cap connections tend to be the first to exhibit insufficient bearing capacity; (4) the critical scour depth according to the original code and the current code is 10.57 m and 0 m, respectively.
Assessing the conditions of masonry arch bridges is a complex challenge that requires proper tools for their survey and structural investigation. This is especially important for most ancient bridges due to their historical value and role in the current infrastructure network. Besides effective analysis methods, minimising computational effort is essential for professional use. The computational costs and analysis time of three-dimensional advanced modelling strategies are poorly compatible with engineering practice. Within this context, the Discrete Macro-Element Method (DMEM) is considered for a detailed study of a historical masonry arch bridge located in northern Portugal, the Barcelos Bridge. After validation in the linear field, pushdown and transversal pushover analyses were carried out, comparing the results with a previous study employing a Finite Element Method (FEM) approach. Moreover, the evaluation of cutwaters modelling in the pushover response and the DMEM mesh sensitivity was considered. The main findings showed a 40% overstrength in the transversal pushover compared to the longitudinal direction due to the cutwaters, which almost double the transversal capacity of the bridge, and a significant reduction in computational costs. Such results support the use of DMEM in professional practice, even with knowledge levels inadequate to resort to advanced FEM models.
Flood behaviour is governed by precipitation and topography, leading to distinct flood properties in mountainous streams and different impact forces on bridge piers. This study develops a hybrid hydrological, hydraulic, and computational fluid dynamics framework to address this issue as an initial step towards comprehensive assessments. First, mountainous streams in Southwest China are examined statistically, revealing that large gradients and sharp bends are typical characteristics. Next, hydrological and hydraulic properties are evaluated with digital elevation models and precipitation depths as inputs. This approach is demonstrated to be capable of assessing discharge for specified rainfall events or for precipitation with prescribed occurrence periods, and to quantitatively estimate the ranges of water depth and velocity along the channel. Subsequently, fluid characteristics accounting for mountainous stream topography are analysed using computational fluid dynamics models, and the resulting impact forces on bridge piers are computed. The steep channel gradients increase the flow velocity and flood forces (i.e. 60% with bed gradients varying from flat to 10%) and generate turbulence downstream of the pier. The velocity, pressure, and vorticity distributions become nonuniform in bent reaches, resulting in complex flow conditions and a 5% variation in the horizontal force on piers.
