Abstract
The structure and properties of grain boundaries in the intermetallic alloy Ni3A1 have received considerable attention in the past several years due to their interesting mechanical properties. The aim in many of these studies [1-13] was to explain the increase in ductility that occurs when polycrystalline Ni3AI is doped with boron. Experimental observations [1-7] have shown that there is a strong correlation between nickel enrichment at grain boundaries and the ductilization of boron-doped polycrystalline Ni3AI. This grain boundary nickel enrichment has only been observed in Nirich Ni3-xAll+x (i.e., x<0). Some experimental studies have shown that grain boundary nickel enrichment also occurs in Ni-rich, boron-free polycrystalline Ni3-xAll+x [3,7]. Two models have been proposed [14,15] to explain the observed ductilization of polycrstalline Ni3AI. In the first model [e.g. 14], an increase in grain boundary cohesive energy has been attributed to nickel enrichment at the grain boundaries, while in the other model [15], it is suggested that the transmittal of slip across the grain boundary becomes easier in chemically disordered grain boundaries, where the disorder is induced by the nickel enrichment at grain boundaries. While these two purported effects of nickel segregation are not necessarily exclusive, experimental studies have been unable to conclusively validate either. There have been several atomistic simulation studies [8-13] that have been carded out to explore this subject. Most of these simulation studies [8-11] have been done at zero temperature and without direct consideration of the important segregation effects, These studies have shown that the cohesive energy for grain boundaries rich in nickel is lower than that for the same boundaries with aluminum-rich or stoichiometric compositions. One of these same studies [9] has shown that the grain boundary cohesive energy can be reduced even further by placing boron atoms in interstitial sites in the boundary, These results seem to give some credibility to the first model, since these changes in the cohesive energy occur for boundaries that are well ordered, One should note, however, that these simulations are carried out with stoichiometric bulk compositions, not at the nickel rich hulk compositions, where nickel enrichment of the grain boundaries and ductilization are experimentally observed. Due to the nature of the zero-temperature, composition simulation method used in these studies, it was not possible to study the variation in grain boundary cohesive energy that occurs with segregation and equilibrium disordering. Monte Carlo simulation studies [12] at I000 K have shown that grain boundaries become enriched with nickel and aluminum for Ni-rich and Al-rich bulk composition, respectively. In this letter, we report the results of our investigations of segregation effects on grain boundary cohesive energy in ordered, boron-fzee Ni3-xAll+x within ~e framework of ~ atomistic simulation procedure. Segregation to two high angle (001) twist grain boundaries ~ (36.9) and 5.13 (22.6) and to (001) free surfaces are studied using a newly developed, free energy minimization method [16-18]. These simulations are carded out for a range of alloy compositions corresponding to nickel concentrations in the bulk of 73.5, 75.0, and 76.6 atomic percent in the temperature range 300-900 K. This model is also used to investigate the degree of chemical disordering that occurs at the grain boundaries. The two existing models f~ the role of Ni segregation on the ductilization of Ni3AI are discussed in light of the present simulation results. Method The local harmonic CL,I-I) model has been applied with considerable success to both perfect and defected single component solids [16,17]. In this model, the classical vibrational contribution to the free energy for a single component system is given by
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have