The cobalt transformation

Abstract

The f.c.c. ⇋ h.c.p. transformation in cobalt powder (equilibrium temperature = 417°C)has been studied by X-ray methods and correlated with the density of stacking faults in both phases. Only the h.c.p. phase is present in the cold-worked powder. The line broadening due to local strains and small particle sizes can be essentially removed by annealing at 300°C, leaving a residual broadening that permits a quantitative determination of the density of random growth and deformation faults. The latter are present in smaller numbers than the growth faults, and are more readily removed by recovery treatments. When the h.c.p. phase is generated from the f.c.c. phase by cooling through the allotropic transformation range, two types of faulted regions become evident: region 1 contains only deformation faults, and region 2 contains both deformation and growth faults. It is suggested that region 1 constitutes the first part of the h.c.p. phase to form on cooling, while region 2 represents the latter part of the transformation. Approximately 25–30 per cent of the parent f.c.c. phase is retained at room temperature under these conditions. Faulting in the f.c.c. phase can be detected with some assurance only after the cooling transformation is underway, but the extent of such faulting is small compared to that in the h.c.p. phase. The observed faulting in both phases is produced mainly by the allotropic transformation, and is not inherited from the parent phase. The faulting generated in the parent phase by the cooling transformation can be partly removed by holding at subcritical temperatures; it is thought that this relaxation process removes barriers to the f.c.c. → h.c.p. transformation and is responsible for the small increments of the h.c.p. phase which form isothermally in the same temperature range. An analysis of the broadening effects in the various diffraction lines suggests that the stacking faults do not terminate within the crystallites or subgrains, but extend to the boundaries. No evidence is found to signify any reversible extension or contraction of the faulting with changing temperature; that is, the faults observed are not in thermodynamic balance.