Continuum modelling of work hardening in precipitation hardened alloys
Time: Mon 2022-12-19 10.00
Location: F3, Lindstedtsvägen 26 & 28, Stockholm
Subject area: Solid Mechanics
Doctoral student: Philip Croné , Hållfasthetslära
Opponent: Professor Norman Fleck, University of Cambridge
Supervisor: Professor Jonas Faleskog, Hållfasthetslära; Professor Peter Gudmundson, Hållfasthetslära
This thesis deals with prediction of macroscopic work hardening in a precipitation hardened alloy. The focus is on the particle contribution. A hierarchical modelling approach is adopted where work hardening in a representative material volume on the microscale is homogenized and used to represent the macroscopic hardening. The modelling on the smaller scale is carried out within the framework of an isotropic continuum strain gradient plasticity theory where particlesare modelled as elastic zones embedded in a continuous isotropic elastic-plasticmatrix. Effects of plastic deformation in smaller particles are included as well.Moreover, the interface between a particle and its surrounding matrix is modelled as a separate region of zero thickness. The end result is an analytical model that highlights the particle contribution under cyclic deformation assuming small plastic strains, and a small to moderate volume fraction of particles. The model moreover allows effects of plastic relaxation around particles to be included in a straightforward manner, which in turn allows larger plastic strains to be considered. Validation of the model is carried out by comparison with experimental uniaxial tension/compression data on a maragin stainless 15-5 steel containingspherical Cu-precipitates. In the first validation, only monotonic loading is considered and the model is brought to close agreement with the data up to a plasticstrain of 7.5% via the implementation of a plastic relaxation model. In the second validation, the model is compared to cyclic tension/compression experiments with plastic strain amplitudes up to 1%. Generally excellent agreement between model and experimental data is obtained.