Modeling of rate dependent deformation of metals and alloys using the viscoplasticity theory based on overstress: Behavior of modified 9chromium-1molybdenum steel at 538 degrees celsius and deformation induced anisotropy at finite strains.

摘要

The modified 9Cr-1Mo steel which is the subject of Part I of this thesis is a relatively new class of steels specifically designed for use in critical high temperature components in the power generation industry. Experimental data of uniaxial loading at 538{dollar}spcirc{dollar}C reveals strong rate dependence, making the Viscoplasticity theory Based on Overstress an ideal starting point model. However, in addition to the more commonly observed properties, the material exhibits permanent softening in the forms of strain rate history dependence of stress levels and cyclic softening, for example. A new history dependent softening formulation is developed to model this behavior. In addition, the 9Cr-1Mo steel exhibits primary, secondary and tertiary creep at quasi-elastic stress levels. A previously formulated recovery of state expression incorporated into the current formulation produces the qualitative simulation of this behavior. The model developed here allows good correlation and predictive abilities for the uniaxial data of the 9Cr-1Mo steel at 538{dollar}spcirc{dollar}C.; Deformation induced anisotropy at finite strains is the subject of Part II of this thesis. When initially isotropic metals are subjected to strains typical of forming processes, they generally develop anisotropic properties. The macroscopic manifestations of induced anisotropy include dimension changes and induced stresses which would not be present in an isotropic metal. The Swift effect seen during torsion is one such example. Plasticity and viscoplasticity models intended to simulate the macroscopic effects of induced anisotropy usually have the same basic structure. They incorporate an isotropic flow rule (and yield criterion if one is used), and an objective backstress growth law, which serves as the repository for induced anisotropy. It will be shown that the popular class of models is incapable of simulating many aspects of the Swift effect. Subsequently, a different class of models is proposed in which induced anisotropy is accounted for in the flow law, which is made to evolve from isotropic to cubic symmetry. Additionally, the characteristic axes of the cubic flow law are allowed to evolve in some cases. The Swift effect can be simulated when the flow law axes are tilted slightly away from the specimen axes. In addition to improved modeling abilities, the proposed class of models can be qualitatively related to induced crystallographic textures which are known to be an underlying cause of much of the anisotropy of interest.