Elisa Bertolesi

Fatigue assessment of ageing riveted railway bridges is challenging due to uncertainties in stress histories and the often-overlooked influence of mean stress on fatigue life predictions. Ignoring mean stress can lead to non-conservative estimates of remaining service life in historic steel structures under variable-amplitude loading. This paper presents a methodology for fatigue evaluation that combines finite element modelling, structural health monitoring (SHM) data, and Eurocode-compliant damage accumulation, explicitly incorporating mean stress correction models. Stress–time histories from numerical simulations are validated against SHM measurements, identifying critical components for detailed fatigue analysis. Fatigue damage is quantified using Miner’s rule, both with and without mean stress correction, to evaluate its impact on service life estimates. A MATLAB-based Fatigue Damage Calculator (FDC) is developed to streamline the assessment, integrating Morrow and Walker mean stress models for systematic comparison. The framework is applied to a historic riveted railway bridge in Alicante, Spain, recently decommissioned. Results show that neglecting mean stress can substantially overestimate fatigue life: uncorrected predictions suggest about 25 years of remaining life, while the Morrow and Walker corrections reduce this to roughly one year or less and indicate potential prior fatigue failure. These findings underscore the importance of incorporating mean stress effects in fatigue assessments of ageing riveted structures. The proposed methodology offers a robust, transferable framework to evaluate structural resilience and inform maintenance and management decisions for heritage railway infrastructure.

Accelerometer and Global Navigation Satellite System (GNSS) can be effectively combined to establish a robust multi-sensor deformation monitoring system. However, GNSS signals may get downgraded in challenging environments, and then destroy the Kalman filter data fusion model. As a result, the accelerometer becomes the only reliable sensor for deformation monitoring, but relying on only accelerometer data may lead to rapid error accumulation due to its potential baseline shift error. To mitigate this challenge, especially in the slow-moving deformation scenarios, we propose a baseline correction prediction algorithm named CNN-BC, based on convolutional neural networks. This algorithm utilizes high-frequency acceleration and baseline correction as input and output features, respectively. The baseline correction of training dataset is derived from the accelerometer and GNSS coupled algorithm. By incorporating the reliable prediction from the network, we can correct the original accelerometer data and reduce error accumulation. To further address the divergence in deformation velocity, we develop CNN-dVel, which uses high-frequency acceleration and velocity difference as input and output features, respectively. We validated the proposed algorithms through two slow deformation experiments utilizing both high-precision and low-cost accelerometers. The results demonstrate that the CNN-BC can predict reliable baseline correction, with an average root mean square (RMS) of 0.37cm/s², and the CNN-dVel achieves non-divergent deformation velocity prediction, with an average RMS of 0.42 cm/s. Furthermore, optimizing the training dataset with acceleration standard deviation (STD) basis enhances prediction accuracy.

Building progressive collapse is currently one of the hottest topics in the structural engineering field. Most of the research carried out to date on this topic has been focused on the structural analysis of the failure of one or more columns in a building to determine the Alternative Load Paths (ALPs) the structure can activate. Past research was mainly focused on extreme situations with high loads and large structural deformations and, to a lesser extent, research looked at lower loads used in design accidental situations, which requires a different set of assumptions in the analysis. This paper describes a study aimed at analysing accidental design situations in corner-column removal scenarios in reinforced concrete (RC) building structures and evaluating the available real ALPs in order to establish practical recommendations for design situations that could be taken into account in future design codes. A wide parametric computational analysis was carried out with advanced Finite Element (FE) models which the authors validated by full-scale tests on a purpose-built building structure. The findings allowed us to: (i) establish design recommendations, (ii) demonstrate the importance of Vierendeel action and (iii) recommend Dynamic Amplification Factors (DAFs) for design situations.