Characterization and micromechanical modelling of microstructural heterogeneity effects on ductile fracture of 6xxx aluminium alloys

Abstract

Ductile failure of three 6xxx serie aluminium alloys has been characterized and modelled for about thirty hardening conditions each. These alloys involve relatively similar composition and volume fraction of second phase particles. The tensile mechanical properties show the expected decrease of fracture strain with increasing strength but also major differences among the different alloys with a factor ten in terms of reduction of area at fracture between best and worst case. The origin of these differences is unraveled by detailed characterization of the void nucleation, growth and coalescence process involving in situ 3D microtomography. A cellular automaton model, involving a high number of particles with distributions of position, sizes and void nucleation stress is developed to predict the fracture strain. Excellent predictions are obtained based on the same unique nucleation stress distribution versus particle size for the three alloys. The key element setting the fracture strain is the effect of particle size distribution and spatial distribution on the void nucleation and coalescence processes. The dependence of ductility on strength is properly captured as well.

Failure of metals I: Brittle and ductile fracture

“…Differences in rate sensitivity between the phases can also impact the void growth rates locally. It is thus not surprising, although generally overlooked in models, that significant differences in void growth rates (factor 5 or more) have recently been documented by tracking in situ the growth of individual voids in 3D tomography; see [275] for Ti and [276] for Al.…”

Numerical validation framework for micromechanical simulations based on synchrotron 3D imaging

“…The calibration of these models is a challenging task that is being addressed in the literature using more advanced experimental and numerical procedures. Some variables such as porosity can now be observed experimentally thanks to X-ray imaging techniques [1,8,69,22] with the possibility of considering inclusions and voids individually based on manual [69] or automatic [22] procedures.…”

“…Further, simulations are necessary to experimentally relate observed quantities such as porosity and number of fractured/debonded inclusions to mechanical variables such as plastic strain and stress-based criteria. These calculations at the microscale are usually conducted with idealistic microstructures and constitutive behavior, considering homogeneous kinematic or static boundary conditions, which cannot capture the local strain and stress states that inclusions and voids are subjected to, due to their random shapes and distribution [1,44,66,22]. The effect of the three-dimensional random distribution of voids on softening has been studied [18] using different void volume fractions.…”

Experimental-Numerical Validation Framework for Micromechanical Simulations

“…These microscale computations are usually driven with idealistic microstructures, constitutive behavior, and simplified kinematic or static boundary conditions that do not capture local strain and stress states that inclusions and voids are subjected to [11,16,17,14]. The principal aim of the present work is to develop reliable simulations at the microscale using validated models to describe the three steps of ductile damage (i.e., nucleation, growth and coalescence).…”

“…Further, the calibration of these models is challenging since they usually require advanced identification techniques [8,9,10]. It is worth noting that some damage variables such as porosity can now be observed experimentally thanks to X-ray imaging techniques [11,12,13,14]. Inclusions and voids can be studied individually based on manual [13,15] or automatic [14] procedures.…”