In recent years, Structural Health Monitoring has become a fundamental area within industrial, aerospace, and civil engineering, offering a wide range of techniques to predict, assess, and manage potential structural damage. Most of the proposed strategies focus on handling and analyzing damage-sensitive features of the system under observation, continuously monitoring and evaluating changes in these characteristics over time. Among the various methods developed recently, vibration analysis has proven to be particularly promising. This work presents an unsupervised learning approach for detecting structural damage using vibration data. The study involves a benchmark test exploiting a custom-designed laboratory truss girder to develop damage detection strategies based on acceleration measurements. The approach employs statistical pattern recognition employing mode-shapes as damage-sensitive features. Distinct attention is given to practical aspects such as sparse sensor layouts, managing long-term monitoring data, and introducing synthetic damage indicators to support maintenance decisions. The algorithm is validated using real data from acceleration measurements on the laboratory truss girder, which has been monitored for several months under varying environmental conditions, including temperature fluctuations and sunlight exposure.

Damage Detection Strategy Based on PCA/Mode-Shapes Developed on a Laboratory Truss Girder Subjected to Environmental Variations

F. Luca';S. Manzoni;
2025-01-01

Abstract

In recent years, Structural Health Monitoring has become a fundamental area within industrial, aerospace, and civil engineering, offering a wide range of techniques to predict, assess, and manage potential structural damage. Most of the proposed strategies focus on handling and analyzing damage-sensitive features of the system under observation, continuously monitoring and evaluating changes in these characteristics over time. Among the various methods developed recently, vibration analysis has proven to be particularly promising. This work presents an unsupervised learning approach for detecting structural damage using vibration data. The study involves a benchmark test exploiting a custom-designed laboratory truss girder to develop damage detection strategies based on acceleration measurements. The approach employs statistical pattern recognition employing mode-shapes as damage-sensitive features. Distinct attention is given to practical aspects such as sparse sensor layouts, managing long-term monitoring data, and introducing synthetic damage indicators to support maintenance decisions. The algorithm is validated using real data from acceleration measurements on the laboratory truss girder, which has been monitored for several months under varying environmental conditions, including temperature fluctuations and sunlight exposure.
2025
Structural Health Monitoring & Machine Learning, Vol. 12 Proceedings of the 43rd IMAC, A Conference and Exposition on Structural Dynamics 2025
File in questo prodotto:
File Dimensione Formato  
RP_9788743801696C18.pdf

Accesso riservato

: Publisher’s version
Dimensione 4.45 MB
Formato Adobe PDF
4.45 MB Adobe PDF   Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11311/1304117
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
  • OpenAlex 1
social impact