Data Pruning of Tomographic Data for the Calibration of Strain Localization Models

Hilth, William; Ryckelynck, David; Menet, Claire

doi:10.3390/mca24010018

Cited by 6 publications

(5 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Without solving this issue, no data oversampling is possible here. Therefore, to limit the memory requirement and computational time, oversampling is coupled with a data pruning technique (Hilth et al, 2019) and numerical approximations. The method intends to restrict Equations ( 2) and (4) to sets of selected rows among the complete set of raws.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Multimodal data augmentation for digital twining assisted by artificial intelligence in mechanics of materials

Aublet

N’Guyen²,

Proudhon³

et al. 2022

Front. Mater.

Self Cite

View full text Add to dashboard Cite

Digital twins in the mechanics of materials usually involve multimodal data in the sense that an instance of a mechanical component has both experimental and simulated data. These simulations aim not only to replicate experimental observations but also to extend the data. Whether spatially, temporally, or functionally, augmentation is needed for various possible uses of the components to improve the predictions of mechanical behavior. Related multimodal data are scarce, high-dimensional and a physics-based causality relation exists between observational and simulated data. We propose a data augmentation scheme coupled with data pruning, in order to limit memory requirements for high-dimensional augmented data. This augmentation is desirable for digital twining assisted by artificial intelligence when performing nonlinear model reduction. Here, data augmentation aims at preserving similarities in terms of the validity domain of reduced digital twins. In this article, we consider a specimen subjected to a mechanical test at high temperature, where the as-manufactured geometry may impact the lifetime of the component. Hence, an instance is represented by a digital twin that includes 3D X-Ray tomography data of the specimen, the related finite element mesh, and the finite element predictions of thermo-mechanical variables at several time steps. There is, thus, for each specimen, geometrical and mechanical information. Multimodal data, which couple different representation modalities together, are hard to collect, and annotating them requires a significant effort. Thus, the analysis of multimodal data generally suffers from the problem of data scarcity. The proposed data augmentation scheme aims at training a recommending system that recognizes a category of data available in a training set that has already been fully analyzed by using high-fidelity models. Such a recommending system enables the use of a ROM-net for fast lifetime assessment via local reduced-order models.

show abstract

Section: Methodsmentioning

confidence: 99%

“…But this dependence is not included in notations here, for the sake of simplicity. The data pruning procedure proposed in Hilth et al (2019) introduces a reduced domain denoted by Ω R . This reduced domain is the support of finite element shape functions related to F:…”

Section: Figurementioning

confidence: 99%

Multimodal data augmentation for digital twining assisted by artificial intelligence in mechanics of materials

Aublet

N’Guyen²,

Proudhon³

et al. 2022

Front. Mater.

Self Cite

View full text Add to dashboard Cite

show abstract

“…In practice, the larger the RID the more accurate the hyper‐reduced prediction when using an approximated reduced basis as a substitute to V ⋆ . Large RID can be obtained by using the k‐SWIM algorithm proposed in Reference . The number of modes N ⋆ is bounded because of the computational complexity of the Newton‐Raphson algorithm applied to solve Equation .…”

Section: Methodsmentioning

confidence: 99%

Hyper‐reduced direct numerical simulation of voids in welded joints via image‐based modeling

Lacourt

Ryckelynck

Forest

et al. 2020

Numerical Meth Engineering

Self Cite

View full text Add to dashboard Cite

Defects such as gas pores can be formed and trapped in the fusion zone during laser welding. These defects can significantly affect the mechanical reliability of the welded joint. Current nondestructive inspection technologies are able to detect micro-voids in a mass production context. Finite element analysis can therefore be used to assess the lifetime of an observed component via image-based modeling. Unfortunately, running a simulation per component entails a huge and generally unaffordable computational cost. In addition, voids do not admit a parametric modeling. In this paper, a numerical method is proposed to study the impact of defects on the mechanical response of a welded joint. It is based on model order reduction techniques that decrease the computational cost of each simulation related to an image-based modeling. To tackle the reduction of nonparametric defects, a multiscale construction of the reduced basis is proposed, although no scale separation is assumed when computing the mechanical response of the structure. Some empirical modes are representing the structure behavior and other empirical modes are related to the defect-induced local fluctuations. They are then assembled to simulate a defective joint. Assets and limitations of the proposed method are explored through a simplified two-dimensional (2D) problem. For the sake of reproducibility, this 2D problem is fully parametric. Finally, a realistic three-dimensional (3D) industrial case is presented, where voids geometries have been measured via computed tomography. This 3D problem being nonparametric, fluctuation modes must be computed on the fly, once the computed tomography has been performed. K E Y W O R D Scombinatorial model order reduction, elasto-plasticity, impossible sampling, material health monitoring, reduced order model INTRODUCTIONDirect numerical simulations (DNSs) have been introduced in fluid mechanics to account for the wide range of scales in turbulent flows, 1 without using a simplified modeling of motions at small scales. In mechanics of heterogeneous Int J Numer Methods Eng. 2020;121:2581-2599. wileyonlinelibrary.com/journal/nme

show abstract

“…• Finite element corrections for displacements and stresses can be easily computed over the RID once the reduced prediction have been achieved. This scheme is termed Hybrid Hyper-Reduction in [ 46]. • A parallel programming of the hyper-reduction method has been proposed in [ 95].…”

Section: Remarksmentioning

confidence: 99%

“…• Reduced order models not only save computational time, they save computational resources including energy consumption savings as explained in [ 90] and memory footprint [ 46].…”

Section: Remarksmentioning

confidence: 99%

Learning Projection-Based Reduced-Order Models

Ryckelynck,

Casenave,

Akkari

2024

SpringerBriefs in Computer Science

Self Cite

View full text Add to dashboard Cite

In this chapter, we introduce the solution space for high-fidelity models based on partial differential equations and the finite element model. The manifold learning approach to model order reduction requires simulated data. Hence, learning projection-based reduced order models (ROM) has two steps: (i) an offline step for the computation of simulated data and for consecutive machine learning tasks, (ii) an online step where the reduced order model is used as a surrogate for the high fidelity model. The offline step generates a train set and a validation set of simulated data. The accuracy and the generalisation of the reduced order model is evaluated in the online step by using a test set of data forecast by the high-fidelity model. The test set aims also to check the computational speedups of the reduced-order model compare to the high-fidelity model.

show abstract

Data Pruning of Tomographic Data for the Calibration of Strain Localization Models

Cited by 6 publications

References 47 publications

Multimodal data augmentation for digital twining assisted by artificial intelligence in mechanics of materials

Multimodal data augmentation for digital twining assisted by artificial intelligence in mechanics of materials

Hyper‐reduced direct numerical simulation of voids in welded joints via image‐based modeling

Learning Projection-Based Reduced-Order Models

Contact Info

Product

Resources

About