A nonlinear simulation of a bi-failure specimen through improved discretisation methods: A validation study

Abstract

A robust and efficient scheme is rendered to elastoplastically study the material nonlinearity of structural components. In this investigation, a specimen manufactured from the aluminium alloy AA6061-T6 is considered. It is mechanically loaded under a uniaxial tensile state and the experimental strain datum is analysed by three-dimensional digital image correlation. Due to specific specimen geometry, complex stress states will occur. However, the specimen yields due to an approximated uniaxial stress state. The obtained remote stress/strain from experimental data is used to validate the computational solutions using advanced discretisation approaches. Therefore, as a preliminary numerical study, the model is simulated through the finite element method formulation. Afterwards, another numerical strategy is adopted – the radial point interpolation method. The Newton–Raphson initial stiffness method is thereby adapted to complete the nonlinear solutions algorithm. Furthermore, the elastoplastic demeanour of aluminium alloys is determined with the von Mises yield criterion, an isotropic hardening rule and an associative flow rule. Obtained computational results fit the experimental digital image correlation solution, which allow to conclude that the proposed meshless methodology is efficient and reliable.