Diffusion and Phase Evolution in Multicomponent Multi-Phase Alloys

TL;DR

This study combines experimental and theoretical approaches to analyze diffusion and phase evolution in Ni-Cr-Al multicomponent alloys, introducing new models and measurement techniques to better understand phase boundary movements and diffusion paths.

Contribution

It presents two novel models for multicomponent, multiphase diffusion and demonstrates their consistency with experimental data, including the first measurement of Kirkendall plane shift in such systems.

Findings

The type 0 interphase boundary moves during diffusion, contrary to previous assumptions.

Single beta-phase zones can grow between two-phase zones when terminal compositions are near phase boundaries.

Numerical errors can cause false local extrema in entropy production distribution, which are corrected by the dual-scale model.

Abstract

Diffusion in ternary, multiphase systems was studied theoretically and experimentally in Ni-Cr-Al system at 1200C. The samples were prepared by the multiple method. It has been shown that the method allows obtaining good quality, planar, reproducible and oxide-free diffusion couples. For the first time the concentration profiles were measured by the wide-line EDS analysis. Two models were applied: 1) multi-phase multi-component model based on the bivelocity method in R1 and 2) dual-scale two-phase model, that includes interdiffusion in matrix (global scale in R1) and the diffusion between matrix and precipitates (local scale in R3). The models were used in numerical experiments in which the diffusion paths, the local entropy production rate and the volume fractions of the phases present in the system were calculated. For the first time the Kirkendall plane shift in two-phase system was computed and compared with experimental findings. It has been shown that both models are consistent. The multi-phase, multi-component model was applied to simulate diffusion and the results were compared with experiment. Experimental and numerical results are in agreement. It has been found that: 1) the type 0 interphase boundary (IB) moves during diffusion, contrary to the Morral theorem, 2) in the case when the terminal compositions of the two-phase alloys are close to the phase boundary then the single beta-phase zone can grow between two-phase zones and diffusion path enters the single-phase region, 3) the diffusion path strongly depends on the shape of the phase boundary line and the accuracy of thermodynamical data is a key factor in modelling and 4) the recently reported occurrence of the local extrema in the distribution of entropy production at the IBs is a result of numerical errors and can be eliminated as is shown in dual-scale two-phase model.