abstract
-
Austenite growth in the intercritical annealing of ternary (Fe-C-Mn) and quaternary (Fe-C-Mn-Si) dual phase steels is studied. The growth process is modelled assuming local equilibrium at austenite/ferrite interfaces and diffusion control. An isothermal anneal starting with a ferrite-pearlite mixture is examined. Growth proceeds in four stages: (1) austenite nucleation from pearlite (which is not examined); (2) attainment of uniform carbon activity within each austenite particle; (3) attainment of uniform carbon activity over large distances in the material (e.g., 100 (mu)m); and (4) austenite growth controlled by diffusion of alloying element(s) in ferrite. The last stage is the only stage where significant alloying element partitioning occurs. This leads to the centre of each austenite particle retaining its initial alloying element concentration while the rim is enriched or depleted in alloying element depending on the value of the alloying element diffusion coefficient. This non-uniform concentration profile is associated with the end of austenite growth and results in a volume fraction greater than the equilibrium volume fraction. Full equilibration of the austenite does not occur in practical time periods. The preceding description applies to all cases treated. Material initially homogeneous with respect to alloying elements was examined: (1) for a planar geometry; (2) for a spherical geometry; and (3) with different size particles. A material initially non-uniform with respect to alloy element (e.g., having an initial sinusoidal fluctuation of alloying element) was examined for the case of: (1) a wavelength comparable with spacing between austenite particles; and (2) longer wavelengths. The assumption that local equilibrium established in short times (100 s) was verified with Scanning Transmission Electron Microscope (STEM) observations. STEM and microprobe analysis data also qualitatively verified predicted Mn concentration profiles for long (50 h) anneals. Volume fractions predicted were in reasonable agreement with observed and published volume fraction data.