Massive-Parent Interphase Boundaries and Their Implications on the Mechanisms of the  →  M Massive Transformation in Ti-Al Alloys P. WANG, D. VEERARAGHAVAN, M. KUMAR, and V.K. VASUDEVAN The massive-parent interphase boundaries associated with the  →  M massive transformation in a Ti-46.5 at. pct Al alloy were studied. Special experiments were performed to arrest the transformation at an early stage. Orientation relationships (ORs) between the  M and parent  (retained as  2 ) phases were determined using electron backscattered diffraction (EBSD) in a scanning electron microscope and by electron diffraction, and the interphase boundaries were characterized by two-beam bright- field/weak-beam dark-field (WBDF) transmission electron microscopy (TEM) and high-resolution transmission electron microscopy (HRTEM). The results reveal that the  m nucleates at grain boundaries generally with a low-index Burgers orientation relation and a coherent interface with one parent grain, but grows into the adjacent grain with a high-index/irrational orientation relation. The growth interfaces between the two phases are generally free of misfit dislocations or other defects and consist of curved parts as well as planar facets, whose macroscopic habit planes are of generally high-index/irrational orientation and deviate substantially from the close-packed planes. On an atomic scale, the growth interfaces are sometimes found to be faceted along {111} planes, as well as along other planes, with closely spaced steps, but are concluded to be incoherent with respect to the parent grain into which growth occurs. The implications of these results on the nucleation and growth mechanisms associated with the  -to- M massive transformation are discussed. In particular, the nature of the interphase boundaries and their relation to whether growth occurs by a ledgewise motion of the interfaces or by continuous growth are addressed. I. INTRODUCTION the subject of many articles in this issue, including the pres- ent one, was more contentious, although some consensus THE massive transformation takes place by diffusional appears to have emerged, as detailed in the anchor article [2] nucleation and growth, during which the change in structure of this issue covering the concluding General Discussion to the product phase occurs without a bulk composition session of MT2000. change and by short-range atomic jumps across the parent- Although there is some general agreement on the nucle- massive boundaries. The “MT 2000” symposium, held dur- ation aspects of the massive transformation, several features ing the ASM/TMS Fall meeting in St. Louis, MO from associated with growth have been under debate. These October 9 to 11, 2000, under the sponsorship of the ASMI include: (1) evidence for growth—whether coherent or inco- Phase Transformations Committee, focused on two central herent; (2) orientation relationships—whether high index issues relating to this transformation: (1) whether the mas- or low index; (3) growth rate—whether crystallographic sive transformation can take place below T 0 within the appro- relationships affect the rapidity of growth; (4) interfacial priate two-phase field, and (2) the role of crystallography, structure—whether coherent/partially coherent or incoher- orientation relations, and the nature of the parent-massive ent; and (5) growth mechanism—whether by the ledge interphase boundaries on the mechanisms of nucleation and mechanism or a random transfer of atoms across interphase growth of this transformation. With regard to the first issue, boundaries. Massalski and co-workers [3–10] concluded that as discussed by Aaronson, [1] in this issue, recent experimen- no well-defined crystallographic relations exist between the tal evidence obtained in the Fe-Ni and Ti-Al systems and parent-massive phases during growth and that the interphase theoretical considerations show definitively that the massive boundaries are incoherent and migrate at high speeds by the transformation can occur below T 0 in two-phase field. The random transfer of atoms across them. On the grounds of second issue, which was debated more extensively and is nucleation theory and calculated and experimental nucle- ation rates, Aaronson and associates [11–15] proposed that many of these boundaries are partially coherent during P. WANG, Postdoctoral Research Associate, is with the Division of Engineering, Brown University, Providence, RI 02912. D. growth. Coherency implies the existence of low-index orien- VEERARAGHAVAN, Staff Scientist, is with Digital DNA Laboratories, tation relations between the parent and product phases. Motorola, Inc., Austin, TX 78251. M. KUMAR, Staff Scientist, is with Although evidence for both the absence and presence of the Lawrence Livermore National Laboratories, Livermore, CA 94550. definite crystallographic relationships between the massive V.K. VASUDEVAN, Professor, is with the Department of Materials Sci- ence and Engineering, University of Cincinnati, Cincinnati, OH 45221- product and parent exist, [3–8,10,13–15] they argued that massive 0012. Contact e-mail: vasudevk@email.uc.edu transformations must always involve special crystallo- This article is based on a presentation made at the symposium entitled graphic relations (between the parent and the product) lead- “The Mechanisms of the Massive Transformation,” a part of the Fall ing to coherent nucleation followed by growth, in which the 2000 TMS Meeting held October 16–19, 2000, in St. Louis, Missouri, under the auspices of the ASMI Phase Transformations Committee. interface changes from coherent to partially coherent and METALLURGICAL AND MATERIALS TRANSACTIONS A U.S. GOVERNMENT WORK VOLUME 33A, AUGUST 2002—2353 NOT PROTECTED BY U.S. COPYRIGHT