The inverse Finite Element Method (iFEM) is a model-based technique able to compute the displacement field of a structure from strain measurements and a geometrical discretization of the same structure. In addition to shape sensing, i.e. the computation of displacements, different Structural Health Monitoring (SHM) applications based on iFEM are available in the literature. These are mainly focused on the detection and localization of the damage, without attempting damage size estimation. The latter can be performed with different approaches such as Artificial Neural Networks, however, a prior knowledge of the different damage scenarios would be required. To overcome this issue, the present research proposes a novel iFEM approach to estimate the damage size without creating any database of damage models. The damage is included in the model to decrease the discrepancies between the physical structure and its model. In particular, the damage size is systematically increased until the strain discrepancy between the experimental measurements from sensors and the numerical strain reconstruction is minimized. The technique is experimentally verified on an aluminum plate subject to fatigue crack propagation. The strain field as input to the iFEM is measured with an Optical Backscatter Reflectometry (OBR) fiber and with Digital Image Correlation (DIC).
Crack Size Estimation with an Inverse Finite Element Model
Oboe, D;Poloni, D;Sbarufatti, C;Giglio, M
2023-01-01
Abstract
The inverse Finite Element Method (iFEM) is a model-based technique able to compute the displacement field of a structure from strain measurements and a geometrical discretization of the same structure. In addition to shape sensing, i.e. the computation of displacements, different Structural Health Monitoring (SHM) applications based on iFEM are available in the literature. These are mainly focused on the detection and localization of the damage, without attempting damage size estimation. The latter can be performed with different approaches such as Artificial Neural Networks, however, a prior knowledge of the different damage scenarios would be required. To overcome this issue, the present research proposes a novel iFEM approach to estimate the damage size without creating any database of damage models. The damage is included in the model to decrease the discrepancies between the physical structure and its model. In particular, the damage size is systematically increased until the strain discrepancy between the experimental measurements from sensors and the numerical strain reconstruction is minimized. The technique is experimentally verified on an aluminum plate subject to fatigue crack propagation. The strain field as input to the iFEM is measured with an Optical Backscatter Reflectometry (OBR) fiber and with Digital Image Correlation (DIC).File | Dimensione | Formato | |
---|---|---|---|
EWSHM 2022_ Crack Size Estimation with an Inverse Finite Element Model.pdf
Accesso riservato
Descrizione: manuscript
:
Publisher’s version
Dimensione
11.86 MB
Formato
Adobe PDF
|
11.86 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.