• PDF / 739,873 Bytes
• 5 Pages / 612 x 792 pts (letter) Page_size
• 48 Downloads / 175 Views

DOWNLOAD

REPORT

---

C.H. Shek Department of Physics and Materials Science, City University of Hong Kong, Kowloon Tong, Hong Kong, People’s Republic of China

H. Yang Physics Department, The Chinese University of Hong Kong, Shatin, N.T., Hong Kong, People’s Republic of China (Received 5 February 2001; accepted 26 February 2001)

Recently, it was demonstrated that Ni99.45B0.55 undergoes grain refinement by either remelting or dynamic nucleation at two different undercoolings. In this work, we found that the grain refinement by remelting can be characterized by a gradual disappearance of the texture of the initially formed dendrites in the transition regime while the grain refinement by dynamic nucleation is characterized by a sudden disappearance of texture.

I. INTRODUCTION

When a metallic melt solidifies at a temperature T, the kinetic crystallization temperature, below its thermodynamic melting temperature T1, the microstructure of the specimen depends strongly on the undercooling defined as ⌬T ⳱ T1 − T. If the melt solidifies at small undercoolings, usually dendritic microstructure will be observed. At high undercoolings, grain-refined microstructures are rather universally observed. Grain refinement was first observed by Walker1 in the solidification of undercooled Ni melts. For a nickel-based alloy, there is a transition from coarse dendritic to fine equiaxed grain structure at a critical undercooling of about 170 K. A number of models2– 4 have been proposed to explain the mechanism of grain refinement, including dynamic nucleation, recrystallization, and remelting of initially formed dendrites. For the alloy Ni70Cu30, grain refinement in the undercooled microstructures was explained by Schwarz et al.5 and Karma6 using a model based on remelting of initially formed dendrites. They proposed that remelting of dendrites occurs when the isothermal period following recalescence exceeds the time for breakup of dendrites. In addition to grain refinement at high undercoolings this model also predicts a grainrefined regime observed at small undercoolings. In fact grain-refined regime at small undercooling has been known in some other alloys.7 On the other hand, 1434

http://journals.cambridge.org

J. Mater. Res., Vol. 16, No. 5, May 2001 Downloaded: 05 Apr 2015

Kui and Leung8 showed that grain refinement in Ni was brought about by dynamic nucleation. The shock wave arose from cavitation created by the volume contraction upon rapid crystallization is the origin of the dynamic nucleation. In our previous work reported in Ref. 9, we added a small amount of boron to nickel to form Ni99.45B0.55 and observed the microstructure evolution as the undercooling increased. It turned out that two different grain refinements appeared at low and at high undercoolings, respectively. At low undercooling, the addition of the small amount of boron introduced a gradually grain-refining regime (84.6 K < ⌬T < 160.0 K) that did not exist in pure nickel. The mechanism of this grain refinement was confirmed to be remelting, similar to that of Ni70Cu30 described ab