Question 6 of 8: Nucleation Mechanisms Across Four Transformations

Question 6: Nucleation Mechanisms Across Four Transformations (20 marks)

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.

6.1 — (a) Recrystallization nucleation

Recrystallization nucleation is heterogeneous, and in fact does not follow classical nucleation theory (there is no $\Delta G^*$ energy barrier of the Question 7(a) type to overcome, because a "nucleus" of strain-free material already effectively exists as a low-dislocation-density subgrain before recrystallization is said to begin). The most likely sites are regions of highest local stored strain-energy heterogeneity: sub-grain boundaries within deformation bands, and especially the highly disoriented, high-dislocation-density regions adjacent to prior grain boundaries and around hard second-phase particles (particle-stimulated nucleation). At these sites, sub-grain coalescence or the bulging of an existing high-angle boundary segment produces a small, strain-free volume with a mobile high-angle boundary already in place; this "nucleus" then grows by strain-driven boundary migration (Question 5(b)) consuming the surrounding deformed matrix, with no separate nucleation barrier to overcome — the rate-limiting step is achieving sufficient local misorientation and boundary mobility, not classical nucleation.

6.2 — (b) Interphase precipitation in HSLA steel

Interphase precipitation is heterogeneous nucleation occurring repeatedly, directly ON the moving austenite/ferrite ($\gamma/\alpha$) transformation interface itself, as it advances during the $\gamma\rightarrow\alpha$ transformation. As the $\alpha/\gamma$ boundary migrates, carbon (rejected from the growing ferrite, which has very low carbon solubility) and the microalloying carbide/nitride former (e.g. Nb, Ti, V) both pile up locally at the moving interface, which is itself a high-diffusivity, high-defect-density heterogeneous nucleation site of very low interfacial-energy penalty (the precipitate nucleates coherently with the newly formed ferrite side). Nucleation occurs in sheets, repeatedly, each time the boundary advances a small increment and then briefly pauses (the precipitation event itself locally retards the boundary, giving a stop-start "interphase" growth mode), producing the characteristic periodic, planar rows of fine carbonitride particles seen paralleling the prior $\gamma/\alpha$ interface positions — a signature microstructure of Nb/Ti/V-microalloyed HSLA steels that provides significant precipitation strengthening.

6.3 — (c) Guinier-Preston zone formation

GP zones form by homogeneous nucleation, assisted critically by excess quenched-in vacancies (Question 4(b), mechanism 2). Because GP zones are fully coherent with the matrix (Question 3(a)), their interfacial energy $\gamma$ is extremely low, which by classical nucleation theory (Question 7(a): $\Delta G^*\propto\gamma^3$) makes the nucleation barrier low enough that nucleation can occur uniformly THROUGHOUT the matrix volume rather than requiring a heterogeneous site — there is no strong energetic preference for grain boundaries, dislocations or other defects the way there is for the equilibrium (higher-interfacial-energy) precipitate. The excess vacancies retained from a rapid quench from the solutionizing temperature provide the fast short-range diffusion path needed for solute (e.g. Cu in Al) to cluster into zones at ageing temperatures well below where thermal-equilibrium vacancy concentrations alone could support the required diffusion rate — hence GP-zone formation is also the process most sensitive to quench rate (Question 4(b)) among all the precipitation stages of Question 3(a).

6.4 — (d) Athermal martensite nucleation

Athermal martensite nucleation is heterogeneous, and unlike every other mechanism in this question, it is NOT thermally activated at all (hence "athermal") — it does not proceed by thermally-assisted overcoming of a $\Delta G^*$ barrier with time at a fixed temperature; instead a fixed population of pre-existing potent nucleation embryos (very small regions of favourable local lattice configuration, dislocation arrangement, or residual strain, already present in the parent austenite from prior processing) becomes able to transform, one by one, purely as a function of UNDERCOOLING below $M_s$. As the temperature is lowered, the chemical driving force $\Delta G_v$ increases in magnitude, which by the Question 7(a) relation ($r^*\propto1/|\Delta G_v|$, $\Delta G^*\propto1/\Delta G_v^2$) progressively lowers the critical nucleus size and activation barrier, so that a progressively larger fraction of the pre-existing embryo population becomes capable of triggering a martensite plate at that instant. Each successful nucleation event triggers essentially instantaneous ($\sim10^{-7}$s), diffusionless, shear-dominated growth of a single plate (not a slow, sustained growth process), so the overall transformed fraction is controlled entirely by HOW MANY embryos have become active at the current temperature (a function of $T$ alone), not by how long the material is held there — the defining signature of "athermal" kinetics, in contrast to the isothermal, time-dependent kinetics of every diffusional nucleation mechanism in parts (a)–(c).