Disorder trapping during crystallization of theB2-ordered NiAl compound

Abstract

Using molecular dynamics simulations, disorder trapping associated with solidification is studied for the (100), (110), and (111) growth directions in the B2 NiAl ordered alloy compound. At the high interface velocities studied we observe pronounced disorder and defect trapping, i.e., the formation of antisite defects and vacancies in the crystal at higher than equilibrium concentrations upon rapid solidification. The vacancies are located primarily on the Ni sublattice and the majority of antisite defects are Ni atoms on the Al sublattice, while the concentration of Al on the Ni sublattice is negligibly small. The defect concentration is found to increase in an approximately linear relationship with increasing the interface velocity. Further there is no significant anisotropy in the defect concentrations for different interface orientations. Our results suggest that the currently available models of disorder trapping should be extended to include both antisite defects and vacancies.

Molecular dynamics study of phonon-mediated thermal transport in a Ni50Al50melt: case analysis of the influence of the process on the kinetics of solidification

“…This potential not only accurately reproduces various properties of crystalline NiAl but also gives a realistic description of the principal properties of the liquid NiAl, even though no liquid data were included in the fitting of the potential. In particular, the results of the MD simulations [8,[17][18][19][41][42][43] with this EAM potential of a number of properties of the liquid Ni 50 Al 50 alloy, including diffusion, were found to be in accordance with experimental data [14][15][16][44][45][46], thus demonstrating a good transferability of this potential for an atomistic simulation of liquid NiAl. The melting temperature of B2-NiAl predicted by this potential is 1520 K [17].…”

“…Now, let us estimate quantitatively the variation of temperature and concentration that can be expected to form near a planar solid-liquid interface in the liquid Ni 50 Al 50 alloy model during simulation of non-equilibrium solidification. Using the maximum value of 3.74 m s −1 of the interface velocity, v int observed at undercooling ΔT ≈ 230 K [19], one can estimate the variation of temperature near a planar solid-liquid interface in the liquid Ni 50 Al 50 alloy model according to Equation (4) as dT int $ 6 K. For this estimation, beside C P % 4:32Nk B and D q ≈ 2.4×10 −7 m 2 s −1 given in Table 1, we also used: W ∼ 1 nm [18,19], T LH ¼ DH m =C P %645 K (DH m %0.24 eV is computed here), and a 1 = 0.8839 and a 2 = 0.6267 [25,26]. Meanwhile, Equation (4) shows that in this case, M 0 , derived from Equation (2), underestimates the actual kinetic growth coefficient M by about 2-3%.…”

“…Although several experimental [14][15][16] and molecular dynamics (MD) [17][18][19] studies have been performed to investigate the growth velocity in a NiAl melt as a function of undercooling up to 200-300 K, the correspondence between experimental and simulation kinetic growth coefficients as well as the interface undercoolings due to a difference in thermal conductivity (and thermal diffusivity) has not been discussed. Indeed, due to the absence of a free-electron contribution, classical MD simulations of metals are expected to underestimate the magnitude of the thermal conductivity, which includes in this case only the phonon-mediated contribution [20][21][22].…”

Deformation mechanism in NiAl single crystals at low temperatures

“…Literature survey shows that Ni vacancy concentration in NiAl compound crystals is at variance due to the specimen's thermal or mechanical treatment history as well as its shape and size, i.e. whether bulk, coarse powder, or thin film [24][25][26][27][28][29]. The vacancy concentration derived in the present work is not inconsistent with that reported by Collins and Sinha [28], who observed that Nivacancy concentration in milled NiAl compounds with compositions very close to stoichiometric 50:50 composition, is of order 1-10 at.%.…”

Disorder trapping by rapidly moving phase interface in an undercooled liquid

“…The disorder trapping was investigated from theoretic perspective by sharp interface models [2,3,6], a diffuse interface model [11] and methods of atomistic simulation [1,12]. In the present work, using an existing approach for fast phase transformations [13], a diffuse interface model results in a system of hyperbolic equations to describe (i) the rapid interface motion and (ii) the transition from ordered to disordered structures formed from undercooled liquids.…”