| Failure of ductile materials such as structural steel is associated with large amount of plastic deformation. Ductile fracture is characterized by three sequential processes; void nucleation, void growth, and void coalescence. Continuum Damage Mechanics (CDM) has emerged as an attractive approach for predicting nucleation and growth of cracks in ductile materials. However, a comprehensive satisfactory CDM model, which can describe the different physical stages of ductile fracture, has not been formulated yet. Most of the work done in this field can be categorized into two main groups; the first is based on the effective or true stress concept, and the second is based on the dilatational model for porous media. Although the second group has significant advantages by using parameters that can be interpreted physically, the effective stress, which captures important elements of damage softening, was not employed. In addition, the coalescence stage has received less attention compared to void nucleation and growth. It is especially important to understand how the voids link together to form micro-cracks, as well as how they link to the main pre-existing crack to control the crack growth process.; In this research, elasto-plastic damage softening models are formulated (unified) by first implementing the effective stress concept into both the modified Gurson-Tvergaard model and the Thomason coalescence criterion, and second, by integrating the modified models into one complete damage softening model. Numerical simulation using the LS-DYNA explicit solver has been carried out utilizing the user defined material (UMAT) capability. The return mapping algorithm has been used as the main numerical tool in updating the stress in the plastic regime. The UMAT FORTRAN subroutines have been created and inserted into the main program to build the LS-DYNA executable file which was then linked with the input files to provide the finite element solutions. Upon implementation of the improved damage softening models, crack nucleation, crack growth and crack path have been predicted for a pre-cracked plate with center hole. The validity of this work is examined by comparison with the experimental data for the crack path available from Theilig and Buchholz (1999). The results show very good agreement between the experimental data and the proposed theoretical-computational damage prediction. Comparisons of crack propagation and crack paths using the damage models have been made.; The pathological behavior of mesh dependence is a major concern in the local damage approach to fracture. Mesh sensitivity analysis has been performed for mesh refinements along the crack path and perpendicular to it. A particular geometric configuration has been used to ensure a specific crack path. The results, in general, show insignificant mesh dependence behavior on crack initiation and propagation. Finally, evaluation of the J-integral for the damage softening models has been accomplished for the purpose of establishing the link between Fracture Mechanics and Damage Mechanics. |
