Ductile Fracture in Dual-Phase Steel: Theoretical, Experimental and Numerical Study

Abstract

The thesis is focused on quasi-static ductile fracture in the low range of stress triaxiality of the dual-phase steel Docol 600DL. The study includes mechanical testing, theoretical aspects of ductile fracture and numerical simulations. Ductile fracture in the low range of stress triaxiality is an interesting topic since experimental data from various studies suggests that the ductility of the material is not only governed by the hydrostatic stressstate, but is also influenced by the deviatoric stress-state.

The thesis consists of four journal articles bound together by a synopsis, where the introduction gives the motivation for the work and a state-of-the-art of the topics examined in this study. After the introduction, the objectives for the study are given followed by a summary of the work, conclusions and suggestions for further work. Finally the four articles are given. The articles are referred to as parts as they give detailed information of the different parts of the study.

In Part I and Part II the experimental side of the study is described. These parts give detailed information regarding the material, test set-ups, specimen geometries and optical measurement procedures. The results from the experiments in form of the minor versus major principal strain as well as the stress triaxiality, the Lode parameter, and the equivalent strain at fracture are presented and used to describe the material’s fracture characteristics. Macroscopically the material displayed typical ductile behaviour with large strains before fracture. This was confirmed by scanning electron microscopy of selected specimens since all showed ductile dimples. Further, the material displayed a significant drop in ductility when exposed to plane-strain loading; an observation that suggests that the material’s fracture properties are Lode dependent.

In Part III several uncoupled criteria for ductile fracture which explicitly take into account the effect of the Lode dependence were assessed. One of the criteria, the modified Mohr-Coulomb criterion, was taken from literature, while two of the criteria, the extended Cockcroft-Latham criterion and the extended Rice-Tracey criterion are augmented versions of existing criteria. The data given in Part I and II were used in evaluating the various criteria and it was found that the proposed extended Cockcroft- Latham criterion managed to give accurate predictions on the equivalent strain at fracture, while a somewhat larger spread was observed for the modified Mohr-Coulomb criterion and the extended Rice-Tracey criterion.

Part IV focuses on numerical modelling of ductile crack propagation. Here the experimental tests from Part I were used in assessing the extended Cockcroft-Latham and modified Mohr-Coulomb criteria on crack propagation. Additionally the effects of using a high-exponent yield surface and material softening were investigated. It was found that the different fracture criteria as well as a change of yield function had little effect on the crack propagation. By including material softening through damage coupling, slant shear fracture as observed in some of the experiments was captured, but this did not alter the global response in form of the force-displacement curves.