PSI - Issue 84

Michele Morici et al. / Procedia Structural Integrity 84 (2026) 89–96

91

Once a response model has been established, anomaly detection requires the definition of reliable thresholds. Given the inherent variability of measured responses, thresholding strategies may be deterministic or probabilistic. Deterministic approaches often rely on control charts and fixed limits derived from baseline data (Wu et al. 2021), whereas probabilistic approaches typically adopt Bayesian formulations or interval estimation techniques to explicitly account for uncertainty (Xu et al. 2022). This paper presents the design and preliminary results of a static SHM system implemented on an operational roadway bridge selected as a case study. The monitoring layout has been conceived to support the investigation of temperature-induced structural effects under real service conditions, with the specific aim of enabling the development and validation of response models for anomaly detection. The bridge has been instrumented with a set of contact sensors providing continuous measurements of temperature, displacements, and rotations. A preliminary analysis of the monitoring data is presented, highlighting correlations between thermal variations and structural responses. These results provide a basis for the development of robust temperature–response models and contribute to improving the interpretation of SHM data for the long-term assessment and maintenance of roadway bridges. 2. Description of the case study and its monitoring system 2.1. The case study The bridge considered in this study has a total length of 66 m, and it consists of two jointless spans, measuring 33.00 m each. The Reinforced Concrete (RC) deck has a 10.30 m width, and it is composed of a 0.22 m thick slab supported by six I-shaped Prestressed Reinforced Concrete (PRC) beams spaced 1.84 m each, whose height is 1.60 m. Additionally, at both ends of each span deck there are two rectangular-shaped transversal stiff beams whose section measures 0.40 m × 1.60 m. The continuity of the deck is ensured at the slab level through a couple of dywidag steel rods located in correspondence of the abutment A and of the pier. The expansion joint is therefore located only in correspondence of the abutment B. For what concerns the restraints, the two mid beams of the deck are supported by bearings fixed only in the lateral direction. All the supports are free in the longitudinal direction. The RC pier has a circular cross-section of 2.5 m diameter. The RC pier cap has a trapezoidal shape of 2.50 m depth and a tapered height that measures 1.45 m at the pier axis and 0.80 m at its extremes. The pier is supported by a 2×2 group of circular piles, having 1.20 m diameter and length that exceeds 20.00 m. Finally, the abutments consist of RC walls each founded on a 3×2 group of circular piles with 1.20 m diameter. The detailed original project of the bridge, including drawings and reports, is available, and a preliminary inspection has been conducted to verify the consistency between the design specifications and the as is condition of the bridge. The second span of the bridge is currently equipped with a SHM system aiming at evaluating its quasi-static behaviour by comprising and, consequently, by deducting the thermal effects, thus leading to a better understanding of the relationship between bridge’s deformation and traffic load. For this purpose, two beams have been equipped with different types of sensors: beam #2 belonging to the slow lane of traffic and beam #5 belonging to the fast one. Moreover, the system is installed on the span #2 between the pier and the abutment B, that are the structural elements whose supports ensure the thermal expansion of the deck. The contact sensors belonging to the SHM system, whose location is schematically depicted in Figure 1, are i ) Linear Variable Displacement transducers (LVDTs); ii ) biaxial tiltmeters; iii ) strain gauge rods; and iv ) thermocouples. Samples are acquired every five minutes. The control of the temperature variations above and under the slab is provided by eleven thermocouples, five of them located over the slab and six under. The thermocouples were inserted at a depth of 7 cm from the external surface to avoid capturing the surface temperature gradient, which is not representative of the actual thermal gradient established in the bridge and could reach locally high values. The deck displacement is monitored through the combination of information provided by vertical and horizontal LVDTs and biaxial tiltmeters. The system includes also strain gauge rods that provide information at local level. In detail, two horizontal LVDTs have been installed at each side of the deck in correspondence of the abutment B. Three vertical LVDTs have been located at mid-span of beams #1, #2 and #5 to provide control on the vertical deflections of both the deck and of the beams at midspan. Five biaxial tiltmeters have 2.2. The case study