21-Mat-A6 Materials Selection and Design for Materials Processing · May 2017
Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)
Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.
Nucleation is HETEROGENEOUS, and essentially never homogeneous. New, strain-free grains form preferentially at pre-existing high-angle grain boundaries, deformation/shear bands and, where present, large second-phase particles (particle-stimulated nucleation). These are the sites of the highest local stored-energy GRADIENT and lattice curvature, so a small volume there can reach a strain-free, favourably misoriented state (typically by subgrain-boundary migration/coalescence, i.e. strain-induced boundary migration, rather than classical nucleation of a new lattice) at a far lower energy cost than nucleating homogeneously in a uniformly strained grain interior, where no pre-existing orientation or boundary advantage exists.
Nucleation is HETEROGENEOUS and REPEATED, occurring directly ON the moving austenite/ferrite ($\gamma/\alpha$) transformation interface as it advances by a ledge (step) mechanism. Microalloy carbide/nitride formers (Nb, V, Ti) segregate to the advancing, semi-coherent $\gamma/\alpha$ interface, which is itself a low-energy heterogeneous nucleation site (far lower interfacial-energy penalty than nucleating a carbide inside either bulk phase). As the interface advances in small increments, a fresh row of fine carbide/nitride particles nucleates at each successive ledge position and is left behind, embedded in the growing ferrite, producing the characteristic periodic sheets ("interphase precipitation rows") that mark the interface's successive resting positions and provide strong precipitation strengthening.
Nucleation is effectively HOMOGENEOUS, made kinetically possible by the excess quenched-in vacancies retained from the solution-treatment quench. Because GP zones are fully coherent with the surrounding matrix, the interfacial-energy contribution to the nucleation barrier is nearly zero (only elastic coherency-strain energy is paid, no broken-bond interfacial energy), so the barrier is low enough for zones to nucleate uniformly throughout the supersaturated matrix without needing a heterogeneous site. The excess vacancy concentration is what makes this fast at low aging temperature in the first place: vacancies dramatically accelerate the otherwise-sluggish substitutional solute diffusion needed to cluster Cu (or other solute) atoms into zones near room temperature.
Nucleation is HETEROGENEOUS and, distinctively, ATHERMAL — it depends only on how far the temperature has fallen below $M_s$, not on how long it is held there. The accepted mechanism is that sub-critical martensite EMBRYOS (arrays of dislocations/faulted regions) already exist in the parent austenite — at pre-existing dislocation pile-ups, stacking faults, or the tips of previously formed martensite plates — and become operative, growing at close to the speed of sound by a diffusionless, glissile (shear) interface, as soon as the chemical free-energy driving force $\Delta G_{\gamma\to\alpha'}$ (which grows monotonically as $T$ falls below the thermodynamic equilibrium temperature $T_0$ and further below $M_s$) exceeds the resistance to interfacial motion at that embryo. Because no diffusion or thermal activation over a time-dependent barrier is involved, the transformed volume fraction is a function of temperature alone (each additional increment of undercooling activates a new population of embryos), which is precisely the operational definition of athermal transformation and the reason martensite fraction cannot be increased by prolonged isothermal holding at a fixed temperature below $M_s$.
