Department of Mechanics: Seminar: Abstract Peerlings 2017

Micromechanics of fracture initiation in multiphase materials

Tom de Geus, Ron Peerlings, Marc Geers

Man-made materials tend to be either strong or ductile. Many applications, however, require strength and ductility. Materials producers attempt to meet these conflicting requirements by developing materials whose microstructure consist of two (or more) phases – often hard, but brittle, particles embedded in a soft and ductile matrix. A prominent example is Dual-Phase steel, which has become an indispensable material in the car industry.

Whereas multiphase microstructures show promising properties, the full potential of this concept to have 'the best of both' cannot yet be exploited because of a lack of understanding of their failure properties. Given their intrinsically complex microstructure, the failure mechanisms of multiphase materials are also complex. Microcracks and microvoids may be nucleated in the hard or soft phase, or at their interface. They subsequently show a tendency to propagate and coalesce via the soft phase, ultimately resulting in the initiation of a macroscopic crack. All of these stages have been observed to depend heavily on the properties of the phases, their volume fractions and their morphology – as well as on the precise, local loading conditions.

The objective our study is to provide some clarity in this overwhelming picture of competing, heavily interdependent candidate mechanisms and paradoxical experimental observations.

We approach the problem at an abstract level, using a highly idealised micromechanical model which includes only the key features of the microstructure, including the key microstructural failure mechanisms. The model assumes the phases to be randomly distributed in a square grid and defines two competing failure mechanisms – one for each phase. Its simplicity serves multiple purposes, as it (i) greatly facilitates the interpretation of results, (ii) allows one to run many (hundreds of) realisations, thus generating statistically meaningful datasets, and (iii) enables a wide range of parameter variations, in order to study the influence of the parameters of the system on the outcome of the competition. Despite its simplicity, the model captures, and helps to explain, the trends predicted by much more complex models as well as observations made in experiments.

Statistical analysis of the simulation results for two-phase microstructures which loosely represent Dual-Phase steel has revealed that the location of the critical damage events, which eventually lead to failure, is largely pre-determined by the microstructural morphology. Damage initiation is found to be promoted by a particular local arrangement of the phases. If several of these features happen to be present in the microstructure at unfavourable relative positions, this leads to a cascade of damage events which link up to form a crack already at a low level of deformation. A second conclusion is that at low stress triaxialities damage is nucleated mainly in the soft phase, whereas at higher triaxialities it nucleates in the hard phase, but then coalesces preferably via the soft phase. This reconciles, rationalises and sharpens observations that were made earlier in the literature for such materials.