PSI - Issue 84

Vincenzo Gattulli et al. / Procedia Structural Integrity 84 (2026) 41–48

42

Keywords: Type your keywords here, separated by semicolons ;

1. Introduction Structural Health Monitoring Systems (SHMSs) are widely used in aerospace, civil and mechanical engineering to deepen understanding of the behavior of complex engineering structures (Crognale et al., 2023). Underground tunnels hosting critical infrastructures require monitoring solutions capable of ensuring structural safety and operational continuity under severe environmental and accessibility constraints (ITA Working Group on Structural Health Monitoring, 2022). This need is particularly critical in underground research facilities, where strict limits on vibration, temperature, humidity, and electromagnetic interference make even minor fluctuations potentially disruptive to experimental measurements. Such environments are characterized by limited accessibility, high safety requirements, and strong sensitivity to microclimate variations, increasing the need for continuous, reliable, and non intrusive monitoring systems. Conventional monitoring approaches based on periodic inspections and isolated sensing systems are often insufficient to capture the interaction between structural response and environmental conditions, due to their limited temporal resolution and weak integration between geometric and sensor data (Farrar and Worden, 2007). The DT paradigm addresses these limitations by integrating geometric models, monitoring data, and analytical tools within a unified framework (Grieves and Vickers, 2017). However, its application to underground infrastructures remains limited and often lacks validation against physics-based models. In this study, a real-time DT is developed for Hall B of the Laboratori Nazionali del Gran Sasso (LNGS) (Rinaldi et al., 2021), combining high-fidelity geometry, multisensor IoT-based monitoring, and near-real-time data processing. The integration of experimental modal analysis with finite element modelling supports the interpretation of structural behaviour and enables preliminary comparison with numerical models and supports condition assessment in complex underground environments. Despite the growing interest in Digital Twin applications for civil infrastructure, existing implementations often remain limited to data aggregation and visualization, with limited integration between real-time monitoring, structural identification and physics-based model validation. This gap is particularly critical in underground environments, where operational constraints hinder controlled testing and limit the availability of high-quality validation data. In this context, the present work proposes an integrated framework that combines multi-source geometric reconstruction, distributed IoT sensing, and operational modal identification techniques within a real-time DT environment. The objective is not yet to present a fully validated predictive Digital Twin, but rather to establish the operational and analytical basis for progressively developing predictive functions, model updating procedures, and anomaly detection strategies. The following sections describe the case study, system architecture, and application to real-time monitoring at LNGS. 2. Digital Twin Architecture for LNGS Hall B Hall B of the Laboratori Nazionali del Gran Sasso (see Fig. 1), one of the largest underground research infrastructures worldwide, has been here used as an example of the proposed methodology for the realization of a DT. The laboratory is located beneath approximately 1,400 m of rock mass in the Gran Sasso massif in central Italy and hosts high-sensitivity physics experiments that require highly stable structural and environmental conditions. Hall B is a large underground gallery characterized by complex geometry, limited accessibility, and long-term exposure to underground environmental effects. Due to its operational role, inspection activities must be carefully planned to avoid interference with ongoing experiments. Strict constraints on vibration, temperature, humidity, and electromagnetic disturbances significantly limit the effectiveness of conventional monitoring approaches. While the underground environment provides advantages for experimental physics, such as attenuation of cosmic radiation and reduced environmental noise, it also imposes stringent requirements on structural stability, as even minor variations in dynamic response may affect measurement accuracy.