Home
Scholarly Works
Multi-scale modeling of decohesion characteristics...
Journal article

Multi-scale modeling of decohesion characteristics of second phase particles from the matrix in uniaxial tension in a high strength aluminum alloy

Abstract

This research investigates stress evolution and plastic deformation characteristics influencing particle–matrix interface decohesion in AA7075-O aluminum sheets. Employing a combined molecular dynamics (MD) - finite element (FE) approach, three interfaces (Al-η, Al-θ, Al-Fe-rich intermetallic) are studied under various loading conditions. Cohesive properties, represented by traction-separation (T-S) curves, are derived using a novel methodology from MD simulations to estimate critical peak traction (t) initiating decohesion and the work of separation (G) at quasistatic strain rates. It is shown that all three interfaces are weaker in shear than in normal loading. The cohesive property estimates from MD simulations are utilized as input in FE-based models with both real and simplified microstructures of AA7075-O sheet material. These models are subjected to large plastic strains, and the decohesion behavior of each particle type are analyzed. It is shown that particle decohesion is a function of inherent cohesive properties, local inter-particle alignment with respect to loading direction, particle morphology and particle size. Interparticle alignment between 0 and 45 degree promote particle decohesion. Larger sized particles with smaller radial dimensions along loading direction aid early decohesion of particles. Decohesion of a particle can also facilitate debonding of neighbouring particles under continued loading. Fe-rich particles have higher likelihood for decohesion due to their weaker interface. The θ precipitates, despite having comparable interface strength as η particles, manifest a tenfold increase in susceptibility to decohesion due to the effect of particle morphology and size.

Authors

Sarmah A; Jain MK

Journal

Engineering Fracture Mechanics, Vol. 301, ,

Publisher

Elsevier

Publication Date

May 2, 2024

DOI

10.1016/j.engfracmech.2024.110013

ISSN

0013-7944

Labels

Contact the Experts team