
Marco Laconi
System identification of railway bridges through train-induced free-decay vibrations.
Rel. Marco Civera, Giulio Ventura, Eleonora Massarelli. Politecnico di Torino, Corso di laurea magistrale in Ingegneria Civile, 2025
![]() |
PDF (Tesi_di_laurea)
- Tesi
Accesso riservato a: Solo utenti staff fino al 10 Ottobre 2028 (data di embargo). Licenza: Creative Commons Attribution Non-commercial No Derivatives. Download (22MB) |
Abstract: |
Structural Health Monitoring (SHM) emerged as one of the most advanced fields in civil engineering, attracting growing interest due to the deteriorating condition of many civil structures and infrastructures. Public awareness has also been raised by recent structural failures, which have drawn attention to the importance of monitoring and safety. Most of these structures were built during the post-World War II economic expansion and, in the absence of adequate maintenance, are now reaching the end of their estimated service life almost simultaneously. This situation highlights the need for strategies that can accurately assess structural conditions with the support of data, while minimising frequency and cost of manned inspections. Operational Modal Analysis (OMA) allows to reconstruct the dynamic properties of structures, and the identified modal parameters can be tracked over time as damage-sensitive features, providing a valuable tool for the assessment of structural conditions. Among civil infrastructures, bridges represent particularly critical elements, since their damage can have severe functional as well as social and economic consequences. They can be classified according to their service class as pedestrian, road, railway, or aqueduct bridges, each with specific constructional features and service requirements. This work focuses on railway bridges, where passenger comfort and safety play a key role. These requirements imply strict limitations on deformability and vibrations, which are generally satisfied through highly rigid structures. However, such high stiffness leads to very low vibration amplitudes, making the application of OMA based on ambient excitations, challenging and often requiring highly sensitive and costly sensors. To overcome this limitation, the present work exploits OMA with free decays after train-induced vibrations, which generate larger amplitudes and allow the use of less sensitive and more affordable accelerometers. Since the external force cannot be measured directly, the focus is placed on the output only, and more specifically, on the free decay portion that follows the excitation and precedes the return to ambient vibration conditions. The main challenge lies in the automatic identification of this time window, of limited duration, within accelerometric records, which include the entire sequence of events from train approach and after its passage. Thus, the contribution of this M.Sc. Thesis consists in the development of a method to automatically extract the free vibration portion from the measured signals and to use it as input for the Eigensystem Realization Algorithm (ERA). The identified poles are then processed through Automatic Operational Modal Analysis (AOMA) techniques to distinguish physical modes from spurious ones. The method has been first tested on a calibrated numerical model and subsequently applied to two real case studies, with experimental data retrieved from field tests. Finally, the results have been compared with those obtained through the Stochastic Subspace Identification (SSI) algorithm, implemented in both MATLAB and as available in the commercial software ARTeMIS, which instead relies on longer ambient vibration datasets. |
---|---|
Relatori: | Marco Civera, Giulio Ventura, Eleonora Massarelli |
Anno accademico: | 2025/26 |
Tipo di pubblicazione: | Elettronica |
Numero di pagine: | 142 |
Soggetti: | |
Corso di laurea: | Corso di laurea magistrale in Ingegneria Civile |
Classe di laurea: | Nuovo ordinamento > Laurea magistrale > LM-23 - INGEGNERIA CIVILE |
Aziende collaboratrici: | NON SPECIFICATO |
URI: | http://webthesis.biblio.polito.it/id/eprint/37173 |
![]() |
Modifica (riservato agli operatori) |