Phase-field models for ductile fracture. Stress triaxiality dependent approaches applied to pipeline steel
Abstract
Ductile fracture is associated with plastic deformation and the development of voids in the material, which leads to a gradual fracture evolution. Ductile fracture in metals is of great importance as it can occur in large engineering structures such as bridges, skyscrapers, aircraft fuselages or pipelines. The formation of a ductile fracture involves complex micromechanical processes that govern the material behaviour.
Models for the prediction and description of ductile fracture need to take these complex material processes into account. The objective of this work is to consider the micromechanical processes and to regularise the fracture model across a damage zone. The focus is on the incorporation of triaxiality effects, i.e. the development of hydrostatic stresses, during plastic yielding when necking takes place. These are relevant for the prediction of ductile fracture behaviour, as the hydrostatic stresses have an inhomogeneous distribution that affects fracture initiation.
This work develops different phase-field models coupled with plasticity that account for the physical processes occurring at ductile fracture and consider a quasi-static crack evolution. Specifically, a phase-field regularisation of the Gurson-Tvergaard-Needleman model is presented, which depends on the porosity evolution in the material and thus accounts for triaxiality effects. In order to gain more flexibility, this regularisation approach is transferred to a general class of local damage models that take triaxiality effects into account. The generalised approach is applied to von Mises plasticity in combination with a phase-field model based on the local Cockcroft-Latham failure criterion. Additionally, a variational phase-field model coupled with modified Cam-Clay plasticity is proposed. This plasticity model includes deviatoric and hydrostatic stresses and the resulting coupled model is a regularised fracture model that considers triaxiality effects for ductile fracture initiation and evolution. Numerical simulations are performed with the finite element program Abaqus in a staggered approach. Small strains as well as finite strains are considered; the simulations are performed on 13Cr steel, which is a specific high-strength steel. Finite element simulations of the aforementioned models require a fine mesh and can therefore only be applied on a small scale. Therefore, a scaling method is needed when considering engineering applications on a larger structural scale. In this context, a pipe with a realistic pressure load for CO2 transport is considered and the running ductile fracture that occurs is analysed. The large-scale pipe is discretised by shell elements. The scaling is achieved by an approach that involves the accurate calibration of the shell elements using the small-scale fracture models.
Based on simulations with different stress states and varying hydrostatic stresses, the models proposed in this work demonstrate the ability to reflect triaxiality effects. The results from the simulations agree with the experimental data. If only the phase-field variable is considered as a damage parameter, the regularising effect of the phase-field leads to mesh-independent solutions in the finite element simulations.
The results underline the importance of considering triaxiality effects for the accurate prediction of ductile fracture behaviour when necking occurs or in the presence of complex geometries such as notches.