SHAPE-MEMORY EFFECT AND RELATED TRANSFORMATION BEHAVIOR IN AN UNAUSAGED FE-NI-CO-TI ALLOY

The shape memory effect and related characteristic features of the forward and reverse martensitic transformations in an unausaged Fe-31Ni-10Co-3Ti (mass%) alloy have been studied. The reversible movement of the austenite-martensite interface and the velocity of the interface motion were investigated by an optical microscope with a heating stage. The substructures in the reversed austenite and retransformed martensite, and dislocation structures of the interface were observed in detail by an electron microscope. The results are as follows. A fairly good shape memory effect is obtained, namely, 60-90% length recovery in the extension test and a perfect shape recovery in the bending test, are observed in spite of the unausaged state. The recovery stress in the shape memory effect is about 60 MPa. The austenite-martensite interface moves reversibly by thermal cycling, as a result, two way shape memory effect is obtained. The forward movement of the interface on the second cooling is quite easy, but the reverse movement of the interface on the second heating is more difficult. The interface velocity in the first reverse transformation is approximately the same as that of an Fe-31Ni-0.3C alloy which produces carbides on heating. A high density of hairpin-shaped dislocations and dislocation loops is observed in the reverse-transformed austenite, which are quite similar to those first observed in disordered Fe-Pt alloys. Straight parallel dislocations with the spacing of 25-40 nm are aligned at the austenite-martensite interface in both cases where the interfaces exist between the retained martensite and the reversed austenite on the first heating, and the retransformed martensite and the virgin austenite on the second cooling. The spacing of these parallel dislocations is about the same as the width of the transformation twins in the fresh martensite. The generation of dislocations in the reversed austenite and the dislocation structures at the interface seem to be reasonably explained by a mechanism proposed by Kajiwara and Owen.