Assisting Digital Image Correlation with fairly-priced, immersed, imagebased spline modeling for data assimilation in mechanics of materials - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2021

Assisting Digital Image Correlation with fairly-priced, immersed, imagebased spline modeling for data assimilation in mechanics of materials

Résumé

The development of volume imaging, such as X-ray tomography, opens up attractive horizons for data assimilation in mechanics of materials. On the one hand, the reconstructed volumetric images that reveal the internal architecture of the material are used to build so-called Digital Image-Based (DIB) models [1,2]. On the other hand, by using in situ testing machines along with sophisticated image- based kinematic measurement tools such as Digital Image Correlation (DIC), it is possible to assess the effects of loading on internal deformation at various scales [3]. It is then tempting to take advantage of such measurements to validate and update the material properties of the DIB models. The typical materials of interest in this study are cellular-type materials which may include, e.g., metallic/polymeric foams, trabecular bones, cell woods, or additive manufacturing materials such as lattice structures. With such microstructures, we are confronted with a paradox: the scale of the constituents is merged with that of the texture, whereas the texture should be defined at a lower scale. This problem has led DIC users to consider elements of very large size compared to the micro- architecture [3], therefore preventing from performing kinematic measurement under the cellular scale. For the first time, we propose a method that breaks this barrier and reduces the resolution despite the absence of texture. It consists of a DIC method assisted with a physically sound weak regularization [4] using a fairly-priced, immersed, elastic B-spline image-based model [2]. This technique introduces a separation of scales above which DIC is dominant (for following the movement of cell boundaries) and below which it is assisted with local elastic image-based modeling. We show that the method can estimate complex local displacement and strain fields with speckle-free low-definition images, even in non-linear regimes such as local buckling or plasticity. The resulting method undoubtedly constitutes a valuable tool for the long-term goal of data assimilation for cellular material at the micro-scale.
Fichier non déposé

Dates et versions

hal-03433947 , version 1 (18-11-2021)

Identifiants

  • HAL Id : hal-03433947 , version 1

Citer

Ali Rouwane, Robin Bouclier, Jean-Charles Passieux, Jean-Noël Périé. Assisting Digital Image Correlation with fairly-priced, immersed, imagebased spline modeling for data assimilation in mechanics of materials. Virtual Isogeometric Analysis (VIGA) 2021, Sep 2021, Lyon, France. ⟨hal-03433947⟩
28 Consultations
1 Téléchargements

Partager

Gmail Facebook X LinkedIn More