Skip to Main content Skip to Navigation
Conference papers

Regularizing FE-mesh-based digital image correlation with splines

Abstract : Digital Image Correlation is a widely used full-field measurement method in experimental mechanics and its use has become essential for data assimilation in solid mechanics. Stereo Digital Image Correlation is a powerful method to perform 3D full-field measurements on 3D-surfaces, thanks to two cameras or more. Before truly starting the experimentation, precisely measuring the shape of the structure is essential. Using a Finite Element (FE) discretization space to express the structure geometry is relevant to easily compare experiment results with conventional (mostly industrial) FE simulations. However, given its roughness, an analysis-suitable FE mesh leads to an ill-posed shape measurement problem, which produces irregular non-realistic shapes. To address this issue, we propose a non-invasive regularization based on a projection of the delicate FE problem onto a smoother spline space. Thus, certain spline properties as their regularity and exact refinement are used, similarly as for performing structural shape optimization. Two approaches are considered: a CAD-based approach, adapted to applications where a CAD description is of interest, and a Free-Form Deformation based approach, convenient to handle any arbitrary (possibly) unstructured FE meshes.
Complete list of metadata
Contributor : Morgane Chapelier <>
Submitted on : Friday, November 13, 2020 - 10:29:35 AM
Last modification on : Friday, November 27, 2020 - 3:05:15 AM


  • HAL Id : hal-03003281, version 1



Morgane Chapelier, Robin Bouclier, Jean-Charles Passieux. Regularizing FE-mesh-based digital image correlation with splines. Virtual Isogeometric Analysis 2020 (VIGA 2020), Aug 2020, Virtual Conference, Canada. ⟨hal-03003281⟩



Record views