Topotaxy might happen when the κ-carbides are very small nanoscale
precipitates. This leads to hardening but κ carbides are cut by the
lattice dislocations in α crystals and κ-α the interface is coherent. It is
seen clearly that κ-carbide in Fig. 1(c) is not in the case. κ will probably
not be cut by the lattice dislocations in α crystals. The κ-α interface is
incoherent due to large lattice misfit (22.1%). Orowan loops will be
formed and cross slip will be enhanced at the κ carbide. The cross slip
may be particularly important for activating secondary slip and the associated
hardening. Orowan loops will create back stresses leading to
further hardening. This causes the accumulation of significant amount
of dislocations around the κ-α interface. α is a soft matrix. κ-Carbide
is a hard inclusion. The high dislocation density around an interface between
high mechanical contrast crystals (α and κ) is favourable for the
crack initiation [22]. The large lamellar κ-carbides leave long and planar
κ-α interface. The nucleated cracks are easier to propagate along a planar
interface due to less energy dissipation than that along a fractal one.
Both the high crack nucleation rate and high crack propagation rate
around κ-α interface lead to a poor formability of Fe-26Mn-9Al-0.75C
steel. κ-γ interface in Fe-34Mn-9Al-0.65C steel, on the contrary, has
small lattice misfit. This causes less high dislocation density and low
crack nucleation rate. The fine structure in higher Mn steel prevents
the nucleated cracks from propagation. This explains why Fe-34Mn-
9Al-0.65C steel has better formability. It has been suggested in literature
that the alloys with coherent interface (i.e. zero misfit, e.g. twining) possesses
ideal mechanical properties [23–24].