Evolution of the microstructure of cobalt during diffusionless transformation cycles

Abstract

Differential scanning calorimetry and transmission electron microscopy have been used to study thermal fatigue due to diffusionless phase transformation cycling in pure cobalt. Thermal cycling through the allotropic (hep ** fee) transformation results in a temperature shift of the calorimetric peaks, which means a delay of the transformation. In addition, the transformation enthalpy, which is greater on heating than on cooling, diminishes when the number of transformation cycles increases. This is interpreted as being due to an evolution of the microstructure. Transmission electron microscopy shows the appearance of transformation-induced defects, which are mainly sessile dislocations. We can interpret the calorimetry results (enthalpy evolution and transformation delay) as due to the interactions between interface dislocations and these sessile dislocations.

Theory of the martensitic transformation in cobalt

“…This was first reported by Adams and Altstetter 65 and confirmed by Munier et al 74 who found a difference ⌬h h Ϫ⌬h c Ӎ5 cal mol Ϫ1 . Both studies also show that when the number of transformation cycles increases ⌬h diminishes and the hysteresis width increases.…”

“…Both studies also show that when the number of transformation cycles increases ⌬h diminishes and the hysteresis width increases. 65,74 A similar interpretation of this ''irreversible'' behavior is proposed by the two groups of searchers: Cycling through the transformation induces defects in the two close packed structures and the difference found for ⌬h h Ϫ⌬h c is due to the amount of energy needed for their formation. The fact that ⌬h h is larger that ⌬h c is consistent with the property assumed in our description, that the fcc phase is more ordered than the hcp phase: More energy will be required to create defects in the fcc phase when heating from hcp, than in the hcp phase, on cooling from fcc.…”

“…Note that there is no contradiction between a decrease of ⌬h and an increase of the region of coexistence of the phases as observed by Munier at al. 74 since ⌬h expresses, in part, the energy required to create new defects the number of which decreases on cycling due to saturation. On the other hand the extension of the hysteresis region should be related to a pinning of an increasing number of defects which favors the phase coexistence.…”

“…Crossing the preceding lines one goes from the starting martensite to the starting austenite points denoted M s and A s in Ref. 74. Further cycling will shift the thermodynamic path towards a larger region of coexistence of the phases.…”

Kinetics of the allotropic hcp–fcc phase transformation in cobalt

“…4, T 0 = 690±7 K (at 1 atm), covers the range of most of the available literature data [2,9,11,18], based on calculations or determined experimentally, taken as average of the "austenite" and "martensite" start temperatures, measured by X-ray diffractometry, magnetometry, dilatometry or calorimetry. Values for the martensite start temperature M S range from 692 K to 661 K and for the austenite start temperature A S range from 694 K to 720 K [2,7,10] , range from 377 to 464 J mol -1 [1,2,9,11,18] and some studies indicate differences upon heating and cooling [10,15]. The diversity of these results appears to depend on the specimen shape (thin layer [6,16,17], powder [3,5,12], sheet [8,10,13,15] and rod [1,3,5,7]), the specimen size [10,15], the state of stress [7,12] and the (type of) thermal treatment [1, 3 -8, 10, 12, 15].…”