The effect of low energy dislocation structures on crack growth onset in brittle crystals

Discontinuous crack growth in semi-brittle crystals is a challenging phenomenon. What are the low energy dislocation structures that accompany stably growing cracks? It has been postulated that shielding dislocations and dislocation-free zones are a result of an equilibrium of forces at the crack tip. In some materials, such as iron, these arguments are used to demonstrate the marginal stability with respect to brittle or ductile fracture. Some recent work on Fe3%Si has shown that applied stress intensities K_I approaching 100 MPa m ^{1 2}. The establishment and re-establishment of under gaseous hydrogen on the {001} cleavage plane. Initial growth starts near a K_I of 20 MPa m ^{1 2}. The establishment and re-establishment of equilibrium with the crack tip dislocation slip band(s) is central to the issues involved in crack stability. For example, is it dislocation shielding by primary emitted dislocations, dislocation cutting and/or shielding as produced by secondary sources, residual stress fields in the wake, the lack of hydrogen or all of the above which leads to crack arrest? To assist in sorting out these complex issues, both crack initiation and arrest in brittle single crystals of Fe3%Si as well as polycrystalline high strength, low alloy steel have been accomplished. Both fracture surface morphology and slip band configurations with and without a prior overload were compared by Nomarski and scanning electron microscopy contrast. Evaluation of dislocation morphologies utilizes backscattered scanning electron microscopy and transmission electron microscopy techniques. The results are rationalized and interpreted in terms of equilibrium solutions for dislocation arrays at a crack tip. This utilizes both superdislocation and array approaches in a computer simulation program incorporating anisotropic elasticity. Equilibrium is simulated using a quasi-static approach. Reports on the relative contribution of dislocation arrays and environmental effects to crack stability are made with the caveat that such simulations are extremely sensitive to the input data set.