Skip to Main content Skip to Navigation
New interface

Ductile failure of random porous materials : Computational approach and application to weld defects

Abstract : During welding, gas bubbles can be trapped within the weld joint. These defects reduce the resistance to ductile fracture, but can be detected by non-destructive controls such as X-ray tomography. However, the direct use of defects in a simulation would lead to prohibitive computation times. The objective of the study is to analyze the influence of pores on the ductile fracture of welded parts, and to propose a method to take into account the tomography images in a simplified way so as to predict efficiently and precisely the resistance in ductile fracture.The influence on ductile fracture is studied by micromechanical simulations on elastoplastic cells containing a random distribution of pores, allowing to represent defect interaction. The study focuses on the detection of fracture at the cell scale either by coalescence or localization, the influence of the loading orientation and the dispersion for different void populations. From simulations on random microstructures, the dependence of the localization strain on the loading conditions can be represented by a metamodel obtained by kriging. A multi-fidelity approach is used to combine results from unit and random cells. Finally, ductile fracture is studied experimentally by means of tests observed by synchrotron tomography, allowing to follow the evolution of defects. The experimental results are compared to the simulation, both at the level of macroscopic behavior, and at the local level of the deformation of defects.
Complete list of metadata
Contributor : ABES STAR :  Contact
Submitted on : Wednesday, November 23, 2022 - 10:38:18 AM
Last modification on : Thursday, November 24, 2022 - 3:54:00 AM


Version validated by the jury (STAR)


  • HAL Id : tel-03867113, version 1



Clément Cadet. Ductile failure of random porous materials : Computational approach and application to weld defects. Materials Science [cond-mat.mtrl-sci]. Université Paris sciences et lettres, 2022. English. ⟨NNT : 2022UPSLM028⟩. ⟨tel-03867113⟩



Record views


Files downloads