Free-Form Deformation Digital Image Correlation (FFD-DIC): A non-invasive spline regularization for arbitrary finite element measurements

Chapelier, Morgane; Bouclier, Robin; Passieux, Jean‐Charles

doi:10.1016/j.cma.2021.113992

Cited by 16 publications

(14 citation statements)

References 67 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Remark Once the algorithm is initialized, we can also benefit from the attractive refinement capabilities of B‐splines to solve the optimization problem (12) in a multilevel fashion (as performed, e.g., for shape optimization 63,64 or in the field of DIC 65,66 ). It consists in first running algorithm (15) using coarse splines spaces for

{\boldsymbol{\lambda}}_c

and

{\boldsymbol{\lambda}}_t

, and then refining the spline spaces (usually by knot‐insertion) and running again algorithm (15) while starting with the solution obtained after the previous optimization.…”

Section: Extending the Vic Methods To Build Image‐based Beam Models O...mentioning

confidence: 99%

Image‐based isogeometric twins of lattices with virtual image correlation for varying cross‐section beams

Passieux

Bouclier

Weeger

2023

Numerical Meth Engineering

Self Cite

View full text Add to dashboard Cite

Cellular materials, in particular metal or polymer foams and beam lattices, are characterized by a complex architecture where the material is concentrated in slender struts. This study is concerned with the construction of numerically efficient sample-specific digital twin for the analysis and characterization of the mechanical behavior of such meta-materials. A method to build an explicit, light, and analysis-suitable isogeometric beam model directly from digital images is proposed. More precisely, as a main contribution, the virtual image correlation method is extended to the case of multiple interconnected beams with varying cross-sections. Special care is also given to the initialization for better efficiency and robustness. Synthetic and real examples of increasing complexity, that is, curved beams, lattices, and foams, are presented to account for the performance of the developed workflow from digital images to an isogeometric twin for mechanical modeling and simulation. K E Y W O R D Scellular materials, digital twin, image-based models, isogeometric analysis, lattices INTRODUCTIONCellular materials are currently gathering an important momentum in both the scientific and industrial communities. 1 They rely upon a simple idea: mimicking nature, such as bones and wood, which allows to achieve, in particular, unprecedented stiffness-to-weight ratios. 2 In practice, cellular materials are characterized by a complex micro-architecture. Most of times, the material is concentrated in small struts connected in different ways, at an intermediate scale between the constituents and the structure. Originally consisting of foams, 3,4 that is, exhibiting a random distribution of cells, the advent of additive manufacturing (AM) now pushes these materials to a new stage: it becomes possible to produce so-called lattices where well-designed unit-cells are periodically (or quasi-periodically 5 ) repeated all over a macro shape to obtain exceptional specific performances. Among others, lattices can be made highly stretchable, 6 auxetic, 7 and multi-functional, and therefore, besides being attractive for lightweight components, they appear of crucial interest for energy absorption, thermal management, design of medical implants, to name a few. [8][9][10][11] However, due to the coexistence of two very different scales, the prediction of the mechanical behavior of such materials remains an issue, especially in non-linear regimes. [12][13][14][15] Their global response is actually intimately related to the local architecture, which may not be perfectly known a priori, either because of the random aspect in foams, or because of the strut-level process-induced defects and uncertainties in lattices. [16][17][18] In view of characterizing the mechanical behavior of such materials, it thus appears of paramount importance to build a digital mechanical twin for each individual tested

show abstract

{\boldsymbol{\lambda}}_c

and

{\boldsymbol{\lambda}}_t

, and then refining the spline spaces (usually by knot‐insertion) and running again algorithm (15) while starting with the solution obtained after the previous optimization.…”

Section: Extending the Vic Methods To Build Image‐based Beam Models O...mentioning

confidence: 99%

Image‐based isogeometric twins of lattices with virtual image correlation for varying cross‐section beams

Passieux

Bouclier

Weeger

2023

Numerical Meth Engineering

Self Cite

View full text Add to dashboard Cite

show abstract

“…There are basically two ways of performing DVC in such situations: strong or weak regularization. The first one aims at reducing the number of unknowns by either increasing the size of the elements [15,25,50] or projecting the problem onto an appropriate reduced basis [51][52][53]. The second approach consists in adding a penalization term to functional (2) in order to improve its convexity [1, 26-28, 37, 38, 54-57].…”

Section: Optimization Schemementioning

confidence: 99%

Architecture-Driven Digital Volume Correlation: Application to the Analysis of In-Situ Crushing of a Polyurethane Foam

et al. 2023

Self Cite

View full text Add to dashboard Cite

Background Digital Volume correlation (DVC) consists in identifying the displacement fields that allow for the best possible registration of volume images of a sample captured at various loading stages. With cellular materials, the use of DVC faces an intrinsic limit: in the absence of an exploitable texture on (or in) the struts or cell walls, the available speckle pattern will unavoidably be formed by the material architecture itself. This leads to the inability of classical DVC techniques to measure kinematics below the cellular scale, i.e. at the sub-cellular or micro scales. Objective Here, we extend a newly developed architecture-driven DIC technique [1] for the measurement of 3D displacement fields in real cellular materials at the scale of the architecture. Methods The proposed solution consists in assisting DVC by a weak elastic regularization using, as support, an automatic finite-element image-based mechanical model. Results Complex (locally buckling) kinematics of a polyurethane foam under compression are accurately measured during an in-situ test. The method is essential to evidence the class of dominance (stretching versus bending) of the foam. Conclusions The proposed method allows to confirm that the foam used is bending-dominated, which is not possible with a classical mesoscopic DVC approach. This method is a good candidate for the analysis of complex local deformation mechanisms at the architecture scale.

show abstract

“…Regularisation strategies must therefore be adopted to circumvent this issue. They usually consist in restricting the subspace in which the shape is sought (whether it be in a strong or a weak sense) (Colantonio et al 2020;Etievant et al 2020;Benning and Burger 2018;Chapelier et al 2021).…”

Section: Intrinsics Extrinsics Shape and Albedo Measurementsmentioning

confidence: 99%

“…From an implementation point of view, we write 𝑈 ext = 𝑁 𝑅 ext 𝑝 ext 0 where 𝑝 ext 0 collects the extrinsics. Shape It is common to measure the shape correction along the normal at the nodes of the mesh (Colantonio et al 2020;Pierré et al 2017;Chapelier et al 2021). Defining the normal at a node is not straightforward, and it is usually done by computing the mean over the normals of neighbouring elements.…”

Section: Extrinsics and Shapementioning

confidence: 99%

Photometric DIC: a unified framework for global Stereo Digital Image Correlation based on the construction of textured digital twins

Fouque¹,

Bouclier²,

Passieux³

et al. 2022

Journal of Theoretical, Computational and Applied Mechanics

View full text Add to dashboard Cite

An innovative approach allowing to rigorously address surface curvature and lighting effects in Digital Image Correlation (DIC) is proposed. We draw inspiration from the research work in Computer Vision (CV) regarding the physical modelling of a camera and adopt it to bring novel and significant capabilities for full-field measurements in experimental solid mechanics. It gives rise to a unified framework for global stereo DIC that we call Photometric DIC (PhDIC). It is based on the irradiance equation that relies on physical considerations and explicit assumptions, which stands for a clear breakthrough compared to the usual grey level conservation assumption. Most importantly, it allows to define a Digital Twin of the Region of Interest, which makes it possible to compare a model with different observations (real images taken from different viewpoints). This results in a consistent formalism throughout the framework, suitable for large-deformation and large-strain displacement measurements. The potential of PhDIC is illustrated on a real case. Multi-view images are first used to measure (or scan) the shape and albedo (sometimes called intrinsic texture) of an open-hole plate. The kinematic basis considered for the displacement measurement is associated to a Finite-Element mesh. Results for the shape and albedo measurement are compared for two completely different sets of pictures. Eventually, a large displacement of the structure is measured using a well-chosen single image.

show abstract

Free-Form Deformation Digital Image Correlation (FFD-DIC): A non-invasive spline regularization for arbitrary finite element measurements

Cited by 16 publications

References 67 publications

Image‐based isogeometric twins of lattices with virtual image correlation for varying cross‐section beams

Image‐based isogeometric twins of lattices with virtual image correlation for varying cross‐section beams

Architecture-Driven Digital Volume Correlation: Application to the Analysis of In-Situ Crushing of a Polyurethane Foam

Photometric DIC: a unified framework for global Stereo Digital Image Correlation based on the construction of textured digital twins

Contact Info

Product

Resources

About