Putting Mechanical Content in DVC: Toward 4D Mechanical Correlation

Abstract : The goal of the present study is to illustrate the full integration of sensor and imaging data into numerical procedures for the purpose of identification of constitutive laws and their validation. The feasibility of such approaches is proven in the context of in situ tests monitored by tomography. The bridging tool consists of spatiotemporal (i.e., 4D) analyses with dedicated (integrated) correlation algorithms. A tensile test on nodular graphite cast iron sample is performed within a lab tomograph. The reconstructed volumes are registered by resorting to integrated digital volume correlation (DVC) that incorporates a finite element modeling of the test, thereby performing a mechanical integration in 4D registration of a series of 3D images. In the present case a non-intrusive procedure is developed in which the 4D sensitivity fields are obtained with a commercial finite element code, allowing for a large versatility in meshing and incorporation of complex constitutive laws. Convergence studies can thus be performed in which the quality of the discretization is controlled both for the simulation and the registration. Incremental DVC analyses are carried out with the scans acquired during the in situ mechanical test. For DVC, the mesh size results from a compromise between measurement uncertainties and its spatial resolution. Conversely, a numerically good mesh may reveal too fine for the considered material microstructure. With the integrated framework proposed herein, 4D registrations can be performed and missing boundary conditions of the reference state as well as mechanical parameters of an elastoplastic constitutive law are determined in fair condition both for DVC and simulation.
Complete list of metadatas

https://hal-upec-upem.archives-ouvertes.fr/hal-01324423
Contributor : Florent Pled <>
Submitted on : Wednesday, June 1, 2016 - 8:21:04 AM
Last modification on : Friday, October 4, 2019 - 1:30:42 AM

File

NCCM_2016_abstract (1).pdf
Files produced by the author(s)

Licence


Public Domain

Identifiers

  • HAL Id : hal-01324423, version 1

Citation

François Hild, Amine Bouterf, Ludovic Chamoin, Hugo Leclerc, Florent Mathieu, et al.. Putting Mechanical Content in DVC: Toward 4D Mechanical Correlation. Worshop New Challenges in Computational Mechanics 2016 (NCCM 2016). 2016. ⟨hal-01324423⟩

Share

Metrics

Record views

404

Files downloads

95