Identification and extraction of damage sensitive features is a crucial task for the health, serviceability, and lifetime assessment of civil and industrial structures. Within this framework, the vibration-based structural damage detection of nonlinear beams is herein accomplished using a convolutional deep autoencoder. First, numerical simulations are performed in which the auto-encoder is trained by exploiting the time history of the response obtained from the nonlinear normal modes of a simply supported nonlinear undamaged beam modelled in COMSOL Multiphysics. Subsequently, the trained autoencoder is used to reconstruct the time history of the response obtained from the nonlinear beam in case of damage. The reconstruction error is examined in order to identify the existence and severity of the damage. These numerical results are also corroborated by laboratory experiments. The experimental setup consists of beam specimens excited by an electrodynamic shaker whereas the dynamic response is acquired on a grid of points using a PSV-3D laser scanning vibrometer. The damage is introduced as a localized reduction of the cross-section of the nonlinear beam.

Damage identification of nonlinear beams based on convolutional deep autoencoder / Joseph, Harrish; Carboni, Biagio; Quaranta, Giuseppe; Lacarbonara, Walter. - (2023). (Intervento presentato al convegno 12th International conference on Structural Dynamics tenutosi a Delft, Netherlands).

Damage identification of nonlinear beams based on convolutional deep autoencoder

harrish joseph
Primo
;
Biagio Carboni
Secondo
;
Giuseppe quaranta
;
Walter Lacarbonara
2023

Abstract

Identification and extraction of damage sensitive features is a crucial task for the health, serviceability, and lifetime assessment of civil and industrial structures. Within this framework, the vibration-based structural damage detection of nonlinear beams is herein accomplished using a convolutional deep autoencoder. First, numerical simulations are performed in which the auto-encoder is trained by exploiting the time history of the response obtained from the nonlinear normal modes of a simply supported nonlinear undamaged beam modelled in COMSOL Multiphysics. Subsequently, the trained autoencoder is used to reconstruct the time history of the response obtained from the nonlinear beam in case of damage. The reconstruction error is examined in order to identify the existence and severity of the damage. These numerical results are also corroborated by laboratory experiments. The experimental setup consists of beam specimens excited by an electrodynamic shaker whereas the dynamic response is acquired on a grid of points using a PSV-3D laser scanning vibrometer. The damage is introduced as a localized reduction of the cross-section of the nonlinear beam.
2023
12th International conference on Structural Dynamics
04 Pubblicazione in atti di convegno::04d Abstract in atti di convegno
Damage identification of nonlinear beams based on convolutional deep autoencoder / Joseph, Harrish; Carboni, Biagio; Quaranta, Giuseppe; Lacarbonara, Walter. - (2023). (Intervento presentato al convegno 12th International conference on Structural Dynamics tenutosi a Delft, Netherlands).
File allegati a questo prodotto
Non ci sono file associati a questo prodotto.

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/11573/1731360
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact