The process kinetics during Transient Liquid Phase (TLP) bonding and other diffusion-controlled, two-phase moving interface problems have been examined using numerical modelling and experimental testing. The review section produces a general background of kinetic modelling of two-phase diffusion-controlled processes. Considerable emphasis is placed on the effect of grain boundary regions on process kinetics, and on previous modelling studies.

Nickel base metals with different grain sizes were TLP-bonded using Ni-19at.%P filler metal. The experiment results produced when bonding single-crystal nickel closely correspond with the calculated output of a one-dimensional model developed in the present thesis. The results of influence of base metal grain boundary regions on the process kinetics form the basis for the subsequent development of two-dimensional finite difference models that accounted for grain boundary-related phenomena.

The one-dimensional, fully implicit finite difference model developed in this thesis permits the calculation of solute distribution and the location of the migrating interface in any diffusion-controlled, two-phase moving interface process. The computed results are in good agreement with experimental results produced in the present thesis, and with results found in the literature (on the solution treatment of α/β brass diffusion couples). The numerical calculation is very fast and accurate. Also, the one-dimensional model permits selection of optimum parameters during TLP-bonding.

The two-dimensional finite difference models developed to examine the effect of grain boundary regions on process kinetics take into account grain boundary diffusion, grain boundary migration and grain boundary grooving (or liquid penetration during TLP-bonding). Modelling has confirmed that the influence of grain boundary regions depends on the grain size, the ratio of grain boundary diffusion and lattice diffusion coefficients. Grain boundary migration only affects the total amount diffused during part of the holding period. The grain boundary grooving model takes into account:​ (1) volume diffusion in each phase, (2) the excess chemical potential resulting from the gradient interfacial curvature, and (3) the excess chemical potential resulting from the balance between the grain boundary energy and interfacial energy at the grain boundary triple junction. This model correctly simulates liquid-solid interface movement at base metal grain boundary regions during TLP-bonding.